HdO.Gorelova.7.vw.L
25-04-2002
15:50
Pagina 1
MANCHU GRAMMAR
HdO.Gorelova.7.vw.L
25-04-2002
15:50
Pagina 2
HANDBOOK OF ORIENTAL STUDIES
HANDBUCH DER ORIENTALISTIK
SECTION EIGHT
CENTRAL ASIA edited by
LILIYA M. GORELOVA
VOLUME SEVEN
MANCHU GRAMMAR
HdO.Gorelova.7.vw.L
25-04-2002
15:50
Pagina 3
MANCHU GRAMMAR
EDITED BY
LILIYA M. GORELOVA
BRILL LEIDEN • BOSTON • KÖLN
2002
HdO.Gorelova.7.vw.L
25-04-2002
15:50
Pagina 4
This book is printed on acid-free paper
Die Deutsche Bibliothek – CIP-Einheitsaufnahme Gorelova, Liliya M.: Manchu Grammar / ed. by Liliya M. Gorelova. – Leiden ; Boston ; Köln : Brill, 2002 (Handbook of oriental studies : Sect.. 8, Central Asia ; 7) ISBN 90–04–12307–5
Library of Congress Cataloging-in-Publication Data Gorelova, Liliya M. Manchu grammar / Liliya M. Gorelova p. cm. — (Handbook of Oriental Studies. Section eight. Central Asia ; vol.7) Includes bibliographical references and index. ISBN 9004123075 (alk. paper) 1. Manchu language—Grammar. I. Gorelova, Liliya M. II. Handbuch der Orientalis tik. Achte Abteilung, Handbook of Uralic studies ; vol.7 PL473 .M36 2002
494’.1—dc21
2001022205
ISSN 0169-8524 ISBN 90 04 12307 5 © Copyright 2002 by Koninklijke Brill NV, Leiden, The Netherlands All rights reserved. No part of this publication may be reproduced, translated, stored in a retrieval system, or transmitted in any form or by any means, electronic, mechanical, photocopying, recording or otherwise, without prior written permission from the publisher. Authorization to photocopy items for internal or personal use is granted by E.J. Brill provided that the appropriate fees are paid directly to The Copyright Clearance Center, 222 Rosewood Drive, Suite 910 Danvers MA 01923, USA. Fees are subject to change. printed in the netherlands
contents
v
CONTENTS
Preface ............................................................................................. xi
Description of Linguistic Materials ................................................ xvii
Part One Prolegomena to Manchu Studies 1. The Altaic Language Family and the Manchu Language ...... 2. The Place of Manchu among the Languages of the Tungus-
Manchu Linguistic Community (from a typological point of
view) ......................................................................................... 3. The Manchus: a Brief Survey of their History ...................... 3.1. The Jurchens and their Ancient History .......................... 3.2. The Manchus’ Ethnicity ................................................... 3.3. Clans in the History of the Manchus ............................... 3.4. Banners and Garrisons as Major Government
Institutions of the Manchus ............................................. 3.5. Manchu Identity ............................................................... 3.6. The Decline of the Manchu Language ........................... 3.7. Rebirth of the Manchus ................................................... 4. The Manchu Language and its Dialects ................................. 4.1. The Manchus and the Sibes ............................................. 4.2. Contemporary Life of the Sibes in Xinjiang, One of the
Most Multinational Regions of China ............................ 4.3. The Sibe Dialect in Relation to Literary Manchu .......... Part Two Manchu Script 1. The History of Writing the Manchu Language ..................... 2. Graphic Structure of the Manchu Letters .............................. 3. The Manchu Graphemes and the Way of their Writing ....... 3.1. The Manchu Graphemes Rendering Vowels (and the
semivowel y) ...................................................................... 3.2. Manchu Graphemes Rendering Consonants .................. 3.3. Additional Graphemes ..................................................... 4. Outward Appearance of Manchu Writing ............................. 5. Punctuation Marks ...................................................................
gorel-vrw.p65
5
4/16/02, 9:34 PM
3
5
6
8
18
19
21
24
27
29
30
31
38
40
49
51
54
55
59
71
72
73
contents
vi
Part Three Phonology 1. Vowels and Consonants ........................................................... 77
1.1. Vowels ............................................................................ 77
1.2. Consonants ....................................................................... 86
2. Vowel Harmony ....................................................................... 89
3. The Dropping of Vowels and Consonants in Manchu Words 99
4. Stress ........................................................................................ 99
5. Types of Syllables and Syllabic Structure of Manchu Words 102
6. Elements of Morphophonemics. Morphemic Structure of a
Word. Stems and Suffixes ........................................................ 112
6.1. Noun Stems ...................................................................... 113
6.2. Verbal Stems ..................................................................... 115
6.3. Alternation of Vowels and Consonants in Stems ............ 116
6.4. Vowel Harmony in Suffixes .............................................. 117
Part Four Morphology 1. Polysemantics and Polyfunctionality of Language Units as
Two Principal Features of Manchu Grammar ....................... 124
2. Nominal Parts of Speech. Different Functional and Semantic
Groups of Nouns ..................................................................... 126
2.1. Human Nouns ................................................................. 133
2.1.1. Plurals .................................................................... 134
2.2. Nouns Denoting Inanimate Objects and Abstract Notions (Object Nouns) ................................................... 140
2.3. Sex of Living Beings ......................................................... 144
2.4. Nouns Denoting Semantics of Quality ............................ 145
2.4.1. The Substantive-Possessive Suffix -ngge .................. 153
2.5. Diminutives and Augmentatives ...................................... 155
2.6. Degrees of Comparison ................................................... 156
2.6.1. Comparative constructions .................................... 157
2.6.2. Superlative Constructions ...................................... 159
2.7. Declension as a Universal System of Syntactic
Subordination. Noun Declension .................................... 162
2.8. Noun Case Forms (Nominative, Genitive, Accusative,
Dative/Locative, Ablative) ............................................... 163
2.9. Word Formation of Nouns ............................................... 194
3. Numerals .................................................................................. 200
3.1. Cardinal Numerals ........................................................... 201
gorel-vrw.p65
6
4/16/02, 9:34 PM
contents
4.
5.
gorel-vrw.p65
vii
3.2. Ordinal Numerals ............................................................ 203
3.3. Distributive Numerals ...................................................... 204
3.4. Fractional Numerals ......................................................... 204
3.5. Multiplicative Numerals .................................................. 205
3.6. Collective Numerals ......................................................... 205
3.7. Iterative Numerals ............................................................ 205
3.8. Numeratives ..................................................................... 206
3.9. Words Denoting Periods of Time. Cycles ....................... 209
3.10. Words for Periods of Time ............................................. 212
Pronouns .................................................................................. 214
4.1. Personal Pronouns. Irregular Pronominal Case Forms ... 214
4.2. Demonstrative Pronouns ................................................. 217
4.3. Possessive Forms of the Personal and Demonstrative
Pronouns ........................................................................... 219
4.4. Interrogative Pronouns .................................................... 219
4.5. Indefinite Pronouns .......................................................... 223
4.6. The Relative Meaning of the Indefinite Pronoun “ele” 224
4.7. Pronominal Use of the Word “beye” .............................. 225
4.8. Nouns in the Pronominal Function ................................. 227
Verbal Morphology ................................................................. 231
5.1. Verbal Parts of Speech ..................................................... 233
5.2. The Structure of a Verbal Stem and Word-Formation
of Verbs ............................................................................ 233
5.3. Verbs of Motion ............................................................... 239
5.4. Aspect .............................................................................. 240
5.5. Voice .............................................................................. 244
5.6. Participles ......................................................................... 252
5.6.1. Imperfect and Perfect Participles .......................... 254
5.6.2. Participles in the Forms of Oblique Cases ........... 257
5.6.3. Declension of Clauses with Participial Predicates 258
5.6.4. Negative Participle Forms ..................................... 260
5.6.5. Substantive Forms of Participles and the Process
of Nominalization of Participles .......................... 263
5.7. Converbs ........................................................................... 267
5.7.1. Imperfect Converbs ............................................... 267
5.7.2. Perfect Converbs .................................................... 276
5.7.3. Conditional Converbs ........................................... 278
5.7.4. Concessive Converbs ............................................. 280
5.7.5. Converbal Forms in -hai/-hei/-hoi ........................... 281
5.7.6. Terminal Converbs ................................................ 282
7
4/16/02, 9:34 PM
contents
viii
6. 7. 8.
9. 10.
5.7.7. Converbal Forms in -nggala/-nggele/-nggolo ............ 283
5.7.8. Negative Forms of Converbs ................................ 284
5.8. Finite Verbal Forms .......................................................... 285
5.8.1. The Indicative. Finite Forms Ascribed to the
Indicative .............................................................. 286
5.8.2. Oblique Moods. Optative. Imperative.
Prohibitive ............................................................ 295
5.8.2.1. Optative .................................................. 295
5.8.2.2. Imperative ............................................... 296
5.8.2.3. Prohibitive ............................................... 304
5.8.3. Forms of Epistemic Modality .............................. 305
5.9. Analytical Verbal Forms and Constructions .................... 309
5.10. Interrogative Forms of Verbs ......................................... 322
Adverbs. Modes of Word-Formation. Different Semantic
Classes of Adverbs ................................................................... 326
Interjections and Onomatopoeic Words ................................. 334
Syntactic Words ....................................................................... 344
8.1 Postpositions ....................................................................... 344
8.2. Conjunctions ..................................................................... 351
8.3. Syntactic Words of Conjunctional Type .......................... 360
Particles .................................................................................... 368
Reduplication of Words. Grammatical Meanings of Pair
Words ....................................................................................... 380
Part Five Syntax 1. Preliminaries ............................................................................ 387
2. Grammatical Means of Expressing Syntactic Relationships
in Manchu .............................................................................. 393
3. The Means of Pragmatic Structuring in Manchu .................. 404
4. Syntax of the Simple Sentence ............................................... 419
4.1. Grammatical Structure of a Nominal Sentence .............. 419
4.2. Grammatical Structure of a Verbal Sentence ................. 436
4.3. Objects (or Complements). Direct and Indirect Objects 453
4.4. Adverbial Modifiers .......................................................... 459
4.5. Attributes .......................................................................... 460
4.6. Homogeneous Parts of the Sentence ............................... 464
5. Syntax of Composite Sentences. Compound and Complex
Sentences .................................................................................. 466
gorel-vrw.p65
8
4/16/02, 9:34 PM
contents
ix
5.1. Compound Sentences ...................................................... 471
5.2. Complex Sentences .......................................................... 472
5.3. Basic Ideas and Terms for the Theory of the Altaic
Hypotaxis ......................................................................... 476
5.4. Altaic Participles as the Basic Non-finite Verbal Forms
Denoting Subordinate Predicativity ................................ 483
5.5. Three Major Classes of Non-finite Verbal Forms Used
to Denote the Subordinate Predication ........................... 507
5.6. Analytical Constructions Functioning as Subordinate
Predicates ......................................................................... 518
5.7. Direct and Indirect Speech .............................................. 526
Postscript .......................................................................................... 530
List of Abbreviations Which Are Used as Grammatical Category
Labels in Interlinear Morphemic Translation ................................ 535
Source Abbreviations ...................................................................... 536
Selected Bibliography ..................................................................... 539
Dictionaries ..................................................................................... 552
Manchu Manuscripts ..................................................................... 552
Archives .......................................................................................... 552
Index of Dialects, Languages and Language Families ................... 557
Index of Names of People, Nations, Places, Countries, Titles,
Periods of Reigns and Proper Names ..................................... 559
Index of Technical Terms .............................................................. 565
Index of Suffixes, Analytical Grammatical Forms and Syntactic
Words ....................................................................................... 593
gorel-vrw.p65
9
4/25/02, 12:06 PM
This page intentionally left blank
contents
xi
PREFACE This Grammar inevitably bears the marks of Russian traditions in Manchu studies. In Russia, Manchu studies began as early as the first quarter of the 18th century. One can, however, safely assume that in the 17th century, interpreters of Manchu, Chinese, and Mongolian must have participated in early frontier diplomatic relations and trade negoti ations, in which Russia and China were engaged. For instance, the Treaty of Nerchin (1689) was drawn up in five languages, namely Manchu, Chinese, Mongolian, Latin, and Russian. The Peking Ecclesiastical Mission, which existed from 1715 until the Treaty of Tianjin and performed secular functions, had played a great role in the development of Manchu studies in Russia. Mem bers and students of the Mission became famous Russian manchurol ogists who created a considerable number of works of great value. Using the Chinese system of teaching Manchu, Russian man churologists were the first scholars who introduced the Manchu language and writing to the Western society. They translated into Russian a number of very important Manchu-Chinese textbooks, which became well known in Western Europe only in the next cen tury. One would not be mistaken in saying that Russian manchurol ogists of the 18th century initiated a distinctive school of Manchu philology, which devoted particular attention to original manuscripts in Manchu. Russian scholars collected these valuable literary mon uments, which in the course of time came to constitute the founda tion of the richest collections of the Asia Department of the Minis try of Foreign Affairs and the Academy of Sciences. They also translated a considerable number of these works into Russian. Unfortunately, the majority of these works still remain unpub lished. Many of them are now kept in the Archives of Orientalists in the Institute for Oriental Studies in the St. Petersburg Branch of the Russian Academy of Sciences. The published ones became an important part of scientific heritage not only in Russia but also outside the country and have not lost their significance to this day. If the first Russian manchurologists, who were excellent connois seurs of the Chinese language, investigated Manchu as a language closely related to Chinese, and may be, to Mongolian, the manchurol
gorel-vrw.p65
11
4/16/02, 9:34 PM
xii
preface
ogists of the beginning of the 20th century began to study the lan guage as one closely related to the Tungusic languages. They de voted particular attention to the establishment and definition of the Tungus-Manchu language family. At the beginning of the 20th century the Manchu language was a specialized field of scientific research at the Oriental institute in Vladivostok (1899-1920). In 1920 the chair of the Manchu language at the Oriental Institute was disestablished. This was connected with the loss of interest in Manchu, which, after the fall of the Qing dynasty (Manchu by origin) in 1911, ceased to function as the offi cial language of China. Subsequently, the Manchu chair was re established at the Oriental faculty of Leningrad (now St. Petersburg) university in 1947, but it existed for only five years. After this period, Russian scholars were chiefly interested in Manchu in regard to its relationship to other languages of the Tungus-Manchu language family. They mainly searched for those language traits, which were lacking for the establishment of the altaicparent language and for the determination of the place of Manchu among other Altaic languages including Tungusic. The necessity to create standard languages and writing systems for Tungusic peoples in the 1930s caused an intensive investigation of all Tungusic languages and dialects scattered over the vast spac es of Siberia and the Far East. During this time a great number of collections of fieldwork texts, textbooks, and dictionaries were made. This was also a time when similarities and distinctions between Manchu and other Tungusic languages were brought under detailed scrutiny. V.A. Avrorin and E.P. Lebedeva who taught me Manchu and some other Tungusic languages at University, were among scholars who participated in a comprehensive project to develop writing systems for the Northern Minority languages in Russia. They were also among those who were involved in the creation of the standard Tungusic languages, wrote scientific grammars of these languages, dictionaries, and textbooks. Doing their fieldwork in Siberia and the Far East, they studied not only the languages, but the history, eth nography, religion, and the way of life of the Tungusic peoples as well. Familiar not only with Manchu, but also (without exaggera tion) with most Tungusic languages, both of them studied the Tungus-Manchu languages, always bearing in mind the similarities be
gorel-vrw.p65
12
4/16/02, 9:34 PM
preface
xiii
tween their grammatical structures. I have tried to follow their ex ample. As it happened, I have been the co-author of the last book, published by E.P. Lebedeva. This is a book dedicated to the lan guage of the Manchu tribe Sibe and partly to the history of these people. The last book on which V.A. Avrorin worked was Manchu Grammar. He didn’t finish the manuscript. For more than 20 years Avrorin’s unfinished book was inaccessible to scholars. Recently it was prepared for publication by A.M. Pevnov and M.M. Khasano va and published in Russia (Avrorin, 2000). The grammar that we suggest to the reader was written before Avrorin’s book came out of press. My first acknowledgment should go to V.A. Avrorin and E.P. Lebedeva who awakened my interest in studying Tungus-Manchu languages. They also introduced me to the prominent Russian tun gusologist, A.F. Boitzova, under whose guidance I wrote my disser tation dedicated to certain verbal categories of Evenki many years ago and who displayed a keen interest in my work until her last days. I am very grateful to her. In the seventies I worked with a group of scholars at the Institute of History, Philology, and Philosophy of the Siberian Branch of the Russian Academy of Sciences. Under the leadership of Prof. Maya Iv. Cheremisina, this group carried out a comparative typological study of complex (polypredicative) sentences in the languages of Siberia. At that time I had a good opportunity to investigate syn tactic structures of different Tungus-Manchu languages trying to find similar and distinct features among them. The scholars of this re search group had shared with me their works in progress, and over the years I greatly benefited from their aid, especially M.Iv. Cher emisina, to whom I owe special thanks. The bulk of the research presented in this Grammar was done at the Department of Linguistics of the Institute for Oriental Studies of the Russian Academy of Sciences. I am sincerely grateful to my colleagues from the Institute for their encouragement and instruc tive discussions. Many of them were of great help to me. Dr. Zoya M. Shalyapina helped me prepare my first articles in English for publication. We spent many hours not only trying to formulate the statements as accurately as possible, but also trying to specify the statements themselves and discussing many linguistic problems. Prof. Evgeniya I. Shutova, a sinologist, taught me Chinese when ten years ago I realized that my work would not be possible without reading books of Chinese specialists in Manchu. I was able to appreciate her
gorel-vrw.p65
13
4/16/02, 9:34 PM
xiv
preface
help when I was reading Manchu grammars recently published in China. I have always been able to consult with Prof. Mariya N. Orlovskaya, a specialist in old and contemporary Mongolian, on any questions concerning Mongolian studies and enjoyed her friendly support over the years. I also enjoyed communicating with Dr. Bulyash H. Todaeva, a specialist in many Mongolian languages, whose books are well-known among specialists. I always had the opportunity to access information about any Altaic publications from Dr. Lev R. Kontzevich, a specialist in Korean, whose encyclopaedic knowledge, provided me with many useful references. I have great ly benefited over the years from the aid of Prof. Vladimir M. Alpa tov, a specialist in Japanese and general linguistics. My special thanks are owed to Dr. Tamara G. Pogibenko who took the trouble of editing the manuscript. The importance of her critical judgment and instructive help can scarcely be exaggerated. I enjoyed working together trying to find better ways of treating specific questions. For many years I have had the opportunity to collaborate with colleagues from the Department of Altaic languages of the Institute of Linguistics of the St. Petersburg branch of the Russian Academy of Sciences. Conferences that were dedicated to Tungus-Manchu studies, and held under the leadership of Prof. Alexander M. Scherbak, always led to further research. Fruitful collaboration with specialists from St. Petersburg, especially with Tatiana A. Pang, has continued until recently. A few words should be said about the extra-linguistic situation at the time when I started my work on Manchu Grammar. For many well-known reasons, scholars in Russia did not have absolute free dom either in getting foreign literature concerning their speciality or in participating in conferences that were conducted outside Russia. These circumstances inevitably caused a lack of information on both sides. Scholars in Russia often had no opportunity to become ac quainted with recent results in their field, which had been obtained in the West, likewise Western scholars did not always know about some of the interesting results obtained in Russia. In my scientific career, the Permanent International Altaistic Conference (PIAC) under the leadership of Prof. Denis Sinor has played a role of great importance. Many of my papers, delivered at the meetings of the Conference, were subsequently published in the Proceedings of the PIAC. It would not be an exaggeration to
gorel-vrw.p65
14
4/16/02, 9:34 PM
preface
xv
say that to a great degree this book is a product of inspiration re ceived from Prof. Denis Sinor. It was he who gave me the idea it self. It was he who during the time of writing and researching in spired me with enthusiasm and encouraged me to further investigation and prevented me from losing courage. Naturally, I would not like him to be too disappointed with the final product. And, of course, he must not be held responsible for any of my in accuracies or mistakes. With the help of the conference, I have had the opportunity to collaborate with my Western colleagues. In 1986, during the meet ing in Tashkent, I made the acquaintance of Prof. Giovanni Stary who has been of very great help to me during many years. He pro vided me with new publications on Manchu studies carried out in China. I have appreciated this greatly not the least because until recently it was rather difficult to get Chinese publications on Man chu studies in Russia. Prof. Stary also published my first articles in English in the “Asi atic Journal” where he has been editor in chief and in a new series “Aetas Manjurica” that has been launched by him together with Martin Gimm and Michael Weiers. I would like to stress that Prof. Stary and the series “Aetas Manjurica” have greatly promoted co operation among scholars engaged in Manchu studies, from every corner of the globe. During many years I enjoyed the help, friendship and support of many other people whom I met at the Conference. I am particularly beholden to Prof. Hans-Peter Vietze who has kindly provided me with the Manchu font. I am very thankful to Dr. Tak-sing Kam from the Hong Kong University of Science and Technology for instructive discussions and useful references he has provided, and to many many others who in different forms provided help, not limited to the writing of this book. Apart from the PIAC, many other institutions and people lent me their assistance and support when my undertaking was just a project. In 1989-1990, I spent half a year in China, at the Liaoning University, and at that time I had contact with Chinese scholars. Specialists in Manchu studies of the Liaoning Academy of Social Sciences were always ready to render assistance in my work. I was very fortunate to meet Prof. Li Shulan and Prof. Hu Zengyi from the Institute of Nationality Studies in Peking and enjoyed fruitful
gorel-vrw.p65
15
4/16/02, 9:34 PM
xvi
preface
discussions with the staff from the Institute of Manchu studies in Harbin, including its former director Prof. Liu Jinsiyan. I am very grateful to colleagues and friends from the University of Auckland in New Zealand, who provided personal and profes sional help. I enjoyed delivering papers at the Linguistic seminar under the leadership of Dr. Frank Lichtenberk. I gratefully acknowl edge the suggestions that I have received in discussing my papers with the members of the Seminar. I am particularly indebted to Prof. Ian Lilly and Dr. Mark Swift, specialists in Russian from the University of Auckland, whose instruc tive help and very useful advice were always of great importance to me during my work. I owe very special thanks to Louise Gauld for helping with my English grammar and style. This book would not have been possi ble without her assistance. She has checked the whole manuscript for readability and ultimately for common sense and I am sorry to say that sometimes it was not very easy to do. I am very thankful to her for all that she has done, beyond any call of duty. I am also very grateful to Sanja Brankov whose diligent editorial work significantly improved the clarity of exposition. At the starting point of my research, from 1996 until 1997, I was subsidized by the Pacific Cultural Foundation of Taiwan. I thank the President of the Foundation Yu-sheng Chang, the Vice-president Nathan Yu-jen Lai, and the Chief of the academic section Margaret Chan with whom I collaborated very fruitfully. As the readers of this book will note, I often give references to the publications of Jerry Norman and Pamela-Kyle Crossley whom I do not know personally but whose books and articles were of great help to me. The readers of this book will perhaps notice that I have not included, for different reasons, references to some publications of Western and Asian specialists. Hopefully, I have compensated for this by bringing to scientific notice a number of works of Russian specialists, which have been previously unknown to Western read ers. And last but not least, I am particularly beholden to my husband Dr. Arkadii M. Slinko whose speciality is far apart from linguistics (although mathematics and structural linguistics are not as far from each other as people used to think), but who, over the years, has shown great patience and rendered his assistance in different ways.
gorel-vrw.p65
16
4/16/02, 9:34 PM
preface
xvii
DESCRIPTION OF LINGUISTIC MATERIALS The linguistic materials used in analyzing different aspects of the Manchu language are taken from several sources. First of all, I have used a number of old Manchu grammars. All of these are listed in Source Abbreviations and Selected Bibliography. Secondly, to a consider able degree, I used material taken from the unpublished manuscript “Sintaks prostogo man’ cz� � urskogo predloz�enija” [Syntax of the Simple Sentence in the Manchu Language] written by B.K. Pashkov. This manuscript is a doctorate dissertation; the author’s abstract of which was published in 1950. The manuscript is kept at the Institute for Oriental Stud ies of the Russian Academy of Sciences (Moscow). It consists of two volumes, the first of which is a description of the structure of the simple sentence in the Manchu language, with examples. The sec ond volume is a collection of specimens from written Manchu, with translations into Russian. This collection includes Manchu texts extracted from books representing various genres of written Manchu. These books can be divided according to the following subject head ings: 1) Manchu-Chinese and Manchu-Mongolian grammars; 2) texts of dialogues; 3) narrative prose; 4) writings of the Chinese wenzhang type. 1) Comparative Manchu-Chinese and Manchu-Mongolian Gram mars (in Manchu, Chinese and Mongolian): a) “Qing wen� qi meng” [Elements of the Manchu Reading and Writing or � The Manchu Language for Beginners] is a title under which a chapter from the earliest Manchu-Chinese dictionary “Dai qing quan shu” (ma. “Dai qing gurun i yooni bithe”) [A Complete Book of the Great Qing Dynasty], published in 1683, was reprinted later on. References to this dictio nary are to be found in the following catalogues: Laufer, 1913:5; Fuchs, 1936:11, 91, 130; Puyraimond, 1979:66-70. A xylograph under the title “Dai qing gurun i yooni bithe” is kept now in the Ar chives of Orientalists of the Institute for Oriental Studies, St. Pe tersburg Branch of the Academy of Sciences of Russia (Volkova, 1988:101-2). This xylograph was described in detail by M.P. Volkova (Volkova, 1974:61-2). The chapter, “Qing wen� qi meng,” was one of the primary standard �
gorel-vrw.p65
17
4/16/02, 9:34 PM
xviii
description of linguistic materials
manuals of Manchu for Chinese students. It was composed by Shou Ping from Changbaishan clan Uge and printed in four volumes (chin. juan) as a xylograph in 1729. This book contained the description of grammatical particles and syntactic words followed by one or two examples in Manchu on every usage. Its manuscript copy from the edition of 1741 is now kept in the Archives of Orientalists of the Institute for Oriental Studies of the St. Petersburg Branch of the Russian Academy of Sciences in the A.O. Ivanovskii’s collection, shifr V-30 (Volkova, 1965:54). The Russian manchurologist I.K. Rossokhin was the first scholar who translated the second part of this book into Russian under the � heading “Osnova man’ cz� urskogo pis’ma” [Elements of the Manchu Writing]” and used it for teaching as a handbook. It should be mentioned that in Europe the first translation of this book from Manchu was car ried out by Alexander Wylie more than a hundred years later, in 1855 (see Bibliography). Fragments of this text were reprinted in the edition “Meadow’s Translations from Manchu, with an Essay on the Lan guage” (1847). As it is well known, Iv.I. Zakharov widely used the text of “Qing wên qi mêng” in his Complete Manchu-Russian Lexicon and Manchu Grammar (Zakharov, 1875:XVI; 1879:II-IV). There are some other descriptions and references to this work in several publica tions of Manchu, such as “Essay on Manchu Literature” by P.G. von Mo�llendorff, “Dictionnaire bibliographique des ouvrage relatifs a l’ Empire � Chinois” by H. Cordier, “Kratkij oc�erk obrazcov man’cz� urskoj literatury” [A Sketch of Specimens of Manchu literature] by A.V. Grebenschikov, “Union catalogue of Manchu books in the National Library of Peking and the Library of the Palace Museum” by Li The Ch’i (Mo�llendorff, 1889-90; Cordier, 1904:2756-7; Grebenschikov, 1909:25; Li, 1933:616.6). P.E. Skachkov presented very interesting information about Russian manchurologists who worked on this particular Manchu text from the beginning to the middle of the nineteenth century, in the book “Ocerki istorii russkogo kitaevedenija” [Essays on the History of Chi � nese Studies in Russia]. He mentions the manuscripts “Analiticeskij razbor � kitaiskich i man’cz� urskich tekstov knigi “Qing wên qi mêng” [Analysis of Manchu � and Chinese texts of the book “Qing wên qi mêng] by O.P. Vojtzekhovskii and “Analiticeskij razbor knigi “Qing wên qi mêng” [Analysis of the book “Qing � wên qi mêng”] by D.A. Peschurov. The first manuscript is probably now kept in the University of Kazan. The second one, which con tains the Manchu text (in Russian transcription) and its Russian translation, is now kept in the Archives of Orientalists of the Insti
gorel-vrw.p65
18
4/16/02, 9:34 PM
description of linguistic materials
xix
tute for Oriental Studies of the St. Petersburg Branch of the Rus sian Academy of Sciences, no. 39, invent. 1, item 1 (Skachkov, 1977:419, 446). The last Russian manchurologist, who worked on translations of “Qing wên qi mêng” into Russian and wrote his own grammatical notes of the Manchu language, which were based on this text, was B.K. Pashkov. However, his manuscript still remains unpublished. It appears that it is now kept at the Institute for Ori ental Studies of the St. Petersburg Branch of the Russian Academy of Sciences. b) “Manju bicihei gisun de aisilara mudan i hergen” (ma.) [Syntactic Words andParticles in the Manchu Written Language], the third volume (chin. juan zhi san; ma. ilaci debtelin) of “Manchu nikan hergen i qing wên qi mêng bithe” (ma.) [Elements of Manchu Reading and Writing in the Manchu and Chinese languages] (Peking 1730, original edition). References to this book are to be found in the following catalogues: Volkova, 1965:54; Ikegami, 1964:114; Li, 1933:616.6; Misig, 1959:194; Mo�llendorff, 1889-90:1; Poppe, 1964:323; Puyraimond, 1979:56; Stary, 1985:38; Fuchs,1936:232. A xylograph under the title “Manchu nikan hergen i qing wên qi mêng bithe” can be found in the Archives of Orientalists of the Institute for Oriental Studies of the St. Petersburg Branch of the Russian Academy of Sciences (Volkova, 1988:123-4). c) “Sula hergen i hacin” (ma.) [Syntactic Words] in: “San he bian lan” (chin.) “Ilan hacin i gisun kamcibuha tuwara de ja obuha bithe” (ma.) [A Book, which Facilitates the Study of Three Languages] (Peking 1792, original edition). d) “Monggo bithei toktoho hergen” (ma.) [The Mongolian Writing]. In: “San he bian lan” (chin.) “Ilan hacin i gisun kamcibuha tuwara de ja obuha bithe” (ma.) [A Book, which Facilitates the Study of Three Languages] (Peking 1792, original edition). The book “Ilan hacin i gisun kamcibuha tuwara de ja obuha bithe” is a manual of Manchu and Mongolian with translation into Chinese and can be found in the following catalogues: Volkova, 1988:1178; Laufer, 1913:3; Li, 1933:613.13; Misig, 1959:101; Poppe, 1964:184; Puyraimond, 1979:77; Stary, 1985:16. It is mentioned in Zakharov’s Complete Lexicon and Manchu Grammar and in “Skizze der manjurischen Literatur” by B. Laufer (Zakharov, 1875:XVIII, 1879:IVV; Laufer, 1927:11). It is also used in “Manchu-mongolische Grammatik” by H.C. Gabelentz (Gabelentz, 1837:255-86). The first publication
gorel-vrw.p65
19
4/16/02, 9:34 PM
xx
description of linguistic materials
of this book is dated 1760. The second one is dated 1792. A xylograph under this title is now kept in the Archives of Orientalists of the Institute for Oriental Studies of the St. Peters burg Branch of the Russian Academy of Sciences (Volkova, 1988:117-8). e)“Dasame foloho manju gisun i untuhun hergen i temgetu jorin bithe” (ma.) [Syntactic Particles in the Manchu Language] (Peking 1896, original edi tion). 2)Books of Manchu-Chinese dialogues:
a) “Tanggu meyen” (ma.) [Hundred Chapters]. In: “Qingwen zhi yao” (chin.) “Manchu gisun i oyonggo jorin i bithe” (ma.) [The Important Manual of the Manchu Language] (Peking 1810, original edition). This book represents a late version of the “Tanggu meyen.” B.K. Pashkov translated Chapters VI-XII of “Tanggu meyen” into Russian. His translations in most part accurately represent the syn tactic structure of Manchu sentences. In several instances Pashkov mentions the existence of alternative versions of the texts, which were used by both himself and Fraser (Fraser, M. & Forbes, A. Tanggu Meyen and Other Manchu Reading Lessons. London, 1924). Sometimes I quote English translations of “Tanggu meyen” taken from the Manchu Grammar by Mo� llendorff who used the texts of “Tanggu� meyen” as reading lessons to elucidate Manchu syntax. As Mo�llendorff men tions, he took English translations from the Chinese version of these dialogues that formed the “Hundred Lessons” of Tomas Wade (Mo�llendorff, 1982:14). References to “Tanggu� meyen” can be found in the following cata logues: Li, 1933:616.8; Misig, 1959:114; Mo�llendorff, 1889-1890:17; Poppe, 1964:321-2; Puyraimond, 1979:57-8; Fuchs, 1966:11, 85. The book is mentioned in Zakharov’s Manchu Grammar (Zakharov, 1879:IV-V). T.A.Pang specially studied “Tanggu� meyen” as a historic and literally monument of the Qing epoch (Pang, 1982:149-155). A xylograph under the title “Manju gisun i oyonggo jorin bithe,” pub lished in 1809, is now kept in the Archives of Orientalists of the Institute for Oriental Studies of the St. Petersburg Branch of the Russian Academy of Sciences (Volkova, 1988:131-2). b) “Dehi meyen” (ma.) [Forty Chapters]. In: “Qing hua wen da si shi tiao dehi meyen” (chin.) (Peking 1759, original edition). B.K. Pashkov trans lated chapters I-V into Russian.
gorel-vrw.p65
20
4/16/02, 9:34 PM
description of linguistic materials
xxi
In the Archives of Orientalists of the Institute for Oriental Stud ies of the St. Petersburg Branch of the Russian Academy of Sciences there is a xylograph under the title “Fan yi si shi tiao” (chin.) “Ubaliyambuha dehi meyen bithe” (ma.) [Forty Chapters, Translated (into Manchu)] (Volkova, 1988:125-6). It was described in detail by T.A. Pang (Pang, 1983:79-84). References to the book “Ubaliyambuha dehi meyen bithe” can also be found in the following catalogues: Mollendorff, 1889-1890:18; Fuchs, 1936:96. c) “Guan hua zhi han” (chin.) [Manchu-Chinese Dialogues of grand style], Chapters I-XIII (the end of the nineteenth century). 3) Narrative prose: a) “Dva brata. Konfucianskij rasskaz [Two Brothers. Confucius tale]. There are not any references, mentioned by B.K. Pashkov. b) “Loo àan alin i doose” [Daos from the Laoshan Mountains]. A story from the collection “Liao zhai zhi yi” (chin.) [The collection of Strange Stories by Liao Zhai] by Pu Sungling. The Chinese writer Pu Sungling wrote under the pen-name Liao Zhai (lived 1640-1715). The whole col lection was finished by Pu Sungling in 1679. The translation of the collected stories into Manchu was published for the first time in 1766. The first publication of the English translation of these stories, made by Giles, was dated 1880, and the third one was dated 1916. The third publication was used by B.K. Pashkov in his article “Liao zhai zhi yi” (Pashkov, 1921:1-24). A xylograph under the title “Sonjofi ubaliyambuha liyoo zhai zhi i bithe” (ma.) [Selected Short Stories by Liao Zhai in translation], dated 1848, is now kept in the Archives of Orientalists of the Institute for Orien tal Studies of the St. Petersburg Branch of the Russian Academy of Sciences (Volkova, 1988:91-2). References to this book can be found in the following catalogues: Li, 1933:707.1.2; Misig, 1959:138; Mo�llendorff, 1889-90:249; Poppe, 1964:524-5; Puyraimond, 1979:131. It was also mentioned by B. Laufer, A.V. Grebenschikov and N. Kanda (Laufer, 1908:15; Grebenschikov, 1909:42-3; Kanda, 1968:70). The publication of the collection “Strange Stories by Liao Zhai”, translated by V.M. Alekseev into Russian, includes the story “Daos from the Mountains of Liao” (Alekseev, 1988:148-65).
gorel-vrw.p65
21
4/16/02, 9:34 PM
xxii
description of linguistic materials
4) Writings of the Chinese wenzhang type: a)“Han i araha mukden i fujurun bithe” (ma.) [Ode to Mukden Writ ten by Khan] (Peking 1748, original edition). This work is well known among western scholars. It was first translated into French by Joseph-Marie Amyot in 1770 and later retranslated and published with the Manchu text by Jules Klaproth in “Chrestomathie Mandchou” in 1828. B.K. Pashkov translated the preface to the “Ode to Mukden” into Russian for his Syntax. See also G. Stary, “L’Ode di Mukden’ dell’imperator Ch’ien-lung” and P. Crossley, “Orphan Warriors.” Crossley writes that much of the language of the Ode was adapted from Umuna’s work (1688) who had been sent to the Northeast by the Kangxi emperor and may ultimately have derived from the work by the Chinese scholar Wu Zhaoqian (Crossley, 1990:235). References to this book can be found in the following catalogues: Li, 1933:702.1; Mo� llendorff, 1889-1890:237; Poppe, 1964:511; Puyraimond, 1979:110; Fuchs, 11936:110. A xylograph under the title “Hani araha mukden i fu fujurun bithe” is now kept in the Archives of Orientalists of the Institute for Orien tal Studies of the St. Petersburg Branch of the Russian Academy of Sciences (Volkova, 1988:99). The text of the book was used by A.V. Grebenschikov in his work “Pis’mennost’ aborigenov Man’ cì� urii” [Writing System of the Native People of Manchuria] (see Grebenschikov, the Archives of Orientalists of the Institute for Oriental Studies of the St. Petersburg Branch of the Russian Academy of Sciences, fond 75, invent. 1, N 35, p.181). b)“Manju gisun i buleku bithei àutucin” (ma.) [Preface to the “Imperial Mir ror Book of the Manchu (Qing) Language]. In: “Qing wen jian” (chin.) (Pe king 1708, original edition). This is a preface, written by the Kangxi emperor (1662-1722), to the explanatory dictionary which was known under the title “Hani araha manju gisuni buleku bithe” (ma.) [Imperial Mirror Book of the Manchu Language]. References to the book “Hani araha manju gisuni buleku bithe” can be found in the following catalogues: Laufer, 1913:41-42; Mo�llendorff, 1889-1890:35; Poppe, 1964:177, 270-4; Puyraimond, 1979:825; Fuchs, 1936:104. The translation of the preface to this book into English was made by A. Fraser (Fraser, 1924:159-65).
gorel-vrw.p65
22
4/16/02, 9:34 PM
description of linguistic materials
xxiii
A xylograph under the title “Hani araha manju gisuni buleku bithe” is now kept in the Archives of Orientalists of the Institute for Orien tal Studies of the St. Petersburg Branch of the Russian Academy of Sciences (Volkova, 1988:102-3). It was studied by M.P. Volkova (Volkova, 1980:74-81). Thirdly, I used as examples on written Manchu the text of the “Niàan samani bithe” (ma.) [The Tale of the Nishan Shamaness] (edition of the book, transliteration from Manchu, translation into Russian and foreword were carried out by M.P. Volkova (see Source Abbre viations and Selected Bibliography). A.V. Grebenschikov found two dif ferent copies of this tale in Manjuria in 1908-1909. The third copy of the tale was gifted to him by a native speaker Manchu, Dekdengge, in 1913. It is this third copy which was translated into Russian by Volkova. As it is well known, this text represents a rare specimen of the original Manchu literature closely related to folk lore. Twenty years ago the Italian manchurologist G. Stary under took a new critical publishing of this text. He used all the known versions and records, which have not been published by M.P. Volkova herself and which are now kept in the Archives of Orientalists at the Institute for Oriental Studies in St. Petersburg. Stary’s publication contains the most complete and informative catalogue of works devoted to “Niàan samani bithe” (Stary, 1985:VIIIX). As Volkova noted, she herself gave the transliteration of the text on the basis of Literary Manchu, therefore the reading of some words of the original manuscript differ from their transliteration (Volkova, 1961:11). As for the syntactic patterns I used from the text of the “Niàan samani bithe” for illustration purposes, they do not differ from those occurring in the Written Manchu language. To illustrate syntactic patterns, I also used the text known as “Sidi Kur,” which was recorded by V.V. Radlov from a native speaker Sibe in 1868-69. Radlov’s linguistic material represented the language of the educated circles of the Sibe community of 1860-70. The edu cated Sibes of the epoch spoke a much purer version of Literary Manchu than the Manchus who settled in Peking and the Jilin and Heilongjiang provinces of China. The “Sidi Kur” was rewritten and translated into Russian by E.P. Lebedeva in 1958-59 (see Source Abbreviations and Selected Bibliography). I assume the right to illustrate my theoretical proposals on Literary (Written) Manchu quoting examples from both Literary Manchu and its Sibe dialect on the
gorel-vrw.p65
23
4/16/02, 9:34 PM
xxiv
description of linguistic materials
grounds of two factors. Firstly, this variant of Sibe, recorded by Radlov, is very close to Literary Manchu; secondly, the syntactic structures that are of interest to me are formed according to basi cally uniform patterns. The main distinctions between Literary Manchu and Radlov’s version of Sibe are of phonetic nature (for more detailed information see Lebedeva & Gorelova, 1994:27-30). In what follows we list the most important phonetic discrepancies: 1. In Sibe the phoneme à occurs more frequently than in Man chu when preceding the vowel i: ma. si—sib. ài = you (2 SG), ma. sikse—sib. àikse “blood,” ma. faksi—sib. fakài “workman,” ma. desi— sib deài “upward.” 2. In Sibe the phoneme v is normally omitted between vowels: ma. kåwaran—sib. kåaran “camp,” “yard,” ma. håwaita-—sib. håaita“to tie,” ma. ferguwecuke—sib. ferguecuke “wonderful,” “astonishing.” 3. The Manchu combination oo (double o) is realized in Sibe as the diphtong [ou]: ma. boo—sib. bou “house,” ma. moo—sib. mou “tree,” ma. hooàan—sib. houàan “paper.” 4. The Sibe vowel u often corresponds to the Manchu vowel o: ma. hoton—sib. hotun “walled city,” ma. oron—sib. orun “vacant post.” In some cases the Sibe vowel o corresponds to the Manchu vowel u: ma. suku—sib. soko “skin,” “pelt,” ma. buta-—sib. bota- “to catch (game or fish).” 5. In Sibe the dropping of vowels in the second syllable of trisyllabic words often occurs: ma. nimaha—sib. nimha “fish,” ma. gicuke— sib. gicku “shameful,” “disgraceful,” ma. fusihån—sib. fuàhån “down,” “downward.” 6. Labial assimilation occurs in several stems in Sibe: ma. kunesun—sib. kunusun “provisions (for journey),” ma. kutele-—sib. kutule“to lead (animals).” At the same time there are several stems in which the reverse correspondence occurs: ma. gucule-—sib. gucele- “to make friends,” ma. kåbule-—sib. kåbili- “to change,” “to become altered.” 7. The Manchu combination ngg is shortened to [�] in Sibe: ma. angga—sib. anga “mouth,” “hole,” ma. ninggun—sib. ningun “six,” ma. erdemungge—sib. erdemunge “virtuous,” “talanted.” In order to facilitate the reader’s understanding of the Manchu language material, I used the form of interlinear morphemic trans lations. Abbreviations used as grammatical category labels, are partly taken from Christian Lehmann, “Directions for Interlinear Morphemic Translations” (see Selected Bibliography).
gorel-vrw.p65
24
4/16/02, 9:34 PM
1
prolegomena to manchu studies
PART ONE
PROLEGOMENA TO MANCHU STUDIES
gorel-1.p65
1
4/17/02, 8:56 AM
This page intentionally left blank
3
prolegomena to manchu studies
PART ONE
PROLEGOMENA TO MANCHU STUDIES 1. The Altaic Language Family and the Manchu Language The beginnings of the Altaic theory date back to the late eighteenth century when linguists started to realize that grammatical structures of many Eurasian languages were characterized by a great number of similarities. Some of these languages had common basic vocabu lary, others displayed interrelations in the lexical sphere. The gram matical system of all these languages could be determined as nominative-accusative. This is to say that structural components of the grammatical system of the language are oriented towards express ing the semantic opposition of subject and object (see also Part 4, Section 2.8). On the lexical level, this opposition manifests itself through the division of all verbs into transitive and intransitive, with pronounced objective orientation of the former, and subjective ori entation of the latter. On the syntactic level, the nominative-accusative system manifests itself through a common nominative struc ture of a sentence which includes a transitive verb-predicate, and through distinguishing the direct and indirect objects as well. The word order of all these languages could be determined as subjectobject-verb type. All of them were agglutinative. This is to say that suffixes denoting various lexical and grammatical meanings may be attached directly to an invariable nominal or verbal stem. Normally a suffix expresses only one meaning, so a string of suffixes can be attached to a stem (we can imagine a train, the locomotive of which is a word stem and the wagons are suffixes). The “Altaic” name was chosen for the theory because the Altaic mountains of Central Asia approximately mark the centre of those lands around which many peoples associated with these languages were located. It is also assumed that peoples speaking the Altaic languages originated in the vicinity of Siberia, and then spread at different times in various directions across Eurasia (Crossley, 1997:34). The Altaic languages can be divided into several groups, prima rily the Turkic, Mongolian, and Tungusic language families. Accord
gorel-1.p65
3
4/17/02, 8:56 AM
4
part one
ing to the latest studies there is evidence that Korean and Japanese are also related to Altaic, but more remotely than the languages from the language families listed above. In scientific literature it is a norm to postulate a more generic language family, viz. the Uralo-Altaic one, which includes the FinnoUgric languages. Among linguistic historians, who have studied the so-called Altaic languages over the past two centuries, there is no complete agree ment on whether or not these languages were genetically related or rather the similarities among them were determined by their ancient contacts. Supporters of the Altaic idea tried from the very begin ning to reconstruct the so-called Proto-Altaic language. To date, a number of versions of sound correspondences among languages referred to as Altaic have been suggested. Some of those hypoth eses were more or less successfully verified. The Manchu language has always been of great interest to spe cialists who devoted much attention to studying the Altaic languages. Manchu could be expected to display features indispensable for the reconstruction of the Proto-Altaic language, which were lacking in other Altaic languages. Being the best-documented member of the Tungusic language family, Manchu has played a significant role in the development and elaboration of the Altaic theory. In accordance with the most ac cepted hypothesis, the Tungusic languages were the first to separate from the Proto-Altaic language, and the Turkic and Mongolian languages were descendants of a younger proto-language (Vladimirt zev, 1929:47). With regard to another hypothesis, the Manchu lan guage deviated significantly from the Proto-Altaic language, and instead of expressing various grammatical meanings with the help of suffixes, as all Altaic languages did, it had developed other ana lytical devices (Grebenschikov, 1912). The term “Altaic” is understood in this work to encompass the typological affinity of the languages. This means that all languages, belonging to the Altaic linguistic community, have to a certain degree, evident similarities in their grammatical structures, in addition to a common basic vocabulary.
gorel-1.p65
4
4/17/02, 8:56 AM
5
prolegomena to manchu studies 2. The Place of Manchu among the Languages of the Tungus-Manchu Linguistic Community (from a typological point of view)
All Tungus-Manchu languages can be divided into three subgroups: the Northern, the Southern and the Manchu-Sibe. The Manchu language and the language of Sibe, a Manchu tribe, because of their peculiarity, form a distinct group. The Northern Tungus-Manchu languages are Even, Evenki, Negidal, and Solon. Peoples speaking these languages inhabit the North of Russia, Siberia and the shores of the Sea of Okhotsk and, to a smaller extent, certain areas of China where one of the Evenki dialects is called Orochen. The Southern Tungus-Manchu languages are Nanai (Gold), Orok, Oroch, Udeghe, and Ulcha. Speakers of these languages inhabit the banks of the Amur river and its tributaries and Sakhalin Island. Other classifications of languages belonging to the TungusManchu language family also exist. For example, J. Ikegami suggested that all Tungus-Manchu languages can be divided into four groups on the grounds of phonetic correspondences existing among the languages. According to him, the first group includes Even, Evenki, Negidal, and Solon. The second group consists of only two languages, Oroch and Udeghe. The third group contains Nanai (Gold), Orok, and Ulcha. The fourth one comprises Manchu and Sibe (Ikegami, 1974:271-2). From the typological point of view the Northern and Southern subgroups of the Tungus-Manchu languages are agglutinative. This means that suffixes displaying various lexical and grammatical meanings are attached directly to stems. Normally a suffix conveys only one meaning, with a few exceptions to the rule, and if it is required to express several language categories a nominal or verbal stem may carry a string of suffixes. The morphological apparatus of all these languages is well developed. For instance, the typical word formation device of Evenki is suffixation. There are more than ten case suffixes, an impressive number of verbal suffixes, including participles (approximately ten suffixes) and converbs (approximately fifteen suffixes). A synthetic language to a great degree, Evenki denotes its grammatical relations within a word form by means of suffixes. The Manchu language and its important dialect, Sibe, because of their particular features, constitute a separate group in terms of
gorel-1.p65
5
4/17/02, 8:56 AM
6
part one
the paradigm of synthetism/analytism. Manchu is the most analyti cal of the languages of the Tungus-Manchu family, and its morpho logical apparatus is only minimally developed. That is, a number of grammatical relationships can be expressed by means of syntac tic words, word order and suprasegmental units. A number of manchurologists have come to the conclusion, through analyzing grammatical forms of the language, that the analytical structure of the language preceded the synthetic one (for instance, B.K. Pashkov in his unpublished manuscript “Syntax of the Simple Sentence in the Manchu Language”). The analytical origin of cer tain Manchu synthetic forms is clear. They are formed by the con traction of two or three stems, which can be easily reconstructed. Analytical elements can be found in various spheres of the structure of the Manchu language: in noun and verb morphology, word for mation and syntax.
3. The Manchus: A Brief Survey of their History According to the population statics given in the third pan-China census, 4,304,981 Manchus and 83,683 Sibes, people of Manchu descent, live in various regions of the People’s Republic of China at present (China. Facts & Figures. Population and Nationalities. Beijing (China), October 1987). In particular they inhabit the provinces of Heilongjiang, Jilin, Liaoning, Hebei and Xinjiang; 117,000 Manchus live in Peking1. In Liaoning province, whose capital city is Shenyang (Mukden), there are eight “Manchu Autonomous Counties” (Xinbin, Feng cheng, Xiuyan, Benxi, Beizhen, Huairen, Kuandian and Qingyuan). In the province of Hebei three “Manchu Autonomous counties” (Qinglong, Fengning, Kuancheng) have been formed. In this prov ince a “Manchu-Mongol Autonomous County,” Weichang, has also been instituted. In Jilin province a “Manchu Autonomous County,” Yitong, has been established. In China over 340 so-called “Manchu Rural Communities” have also been established in recent years (Stary,1995:71). 1
A less reliable figure for the Sibe population, of 172,847 people, is given in two journal publications [see Minzu yanjiu, no. 6, 1984, p.72 and Minzu yanjiu, no. 1, 1991, p.113]. In the opinion of some scholars the difference between the two figures can be accounted for in several ways including a deliberate distortion for political reasons.
gorel-1.p65
6
4/17/02, 8:56 AM
7
prolegomena to manchu studies
In 1982 the Italian scholar A. Pozzi visited Xinbin, the place of origin of the Manchu people, and gathered evidence of the social and linguistic situation that prevailed in the region at that time. The area is situated a few kilometres northeast of Shenyang in Fushun Municipality, and is characterized by its diversity of ethnic nation alities. At the time the Manchus only made up 30 percent of the local population and coexisted with the Mongols, Sibes, Koreans, Hui, and Han peoples. The old Manchu language and culture have been assimilated by the Chinese civilization and no more than a few quite elderly people could still speak Manchu and no one could write it correctly (Pozzi, 1987:208-18). Likewise, in Heilongjiang province a few settlements and villages still remain where some Manchus, again mostly aged people, can still speak their native language. In 1961-62 a number of linguistic expeditions were sent out to these villages of Heilongjiang province, but the results of this activity were only published in 1984. In that area there were four Manchu villages, with a total population of 20,000, only a half of which was ethnically Manchu. In the village of Daujiazi all Manchus over the age of fifty still spoke Manchu. In 1982, when a subsequent expedition took place, almost all Manchus had turned to Chinese and only some seventy-year-old men could speak their native language. In 1961 the Manchu population of another Heilongjiang village, Sanjiazi (in Manchu, “Ilan booi gaàan”— the “Village of the Three Families”), consisted of 355 people and only 58 people, aged over fifty, had excellent knowledge of Manchu; 108 people, aged twenty to fifty, were bilingual (Stary, 1994:476). It is interesting to note that the Chinese scholar An Jun believes that the dialect of the village of Sanjiazi, to the North of Qiqihar, is close to that spoken by the Xinjiang Sibe (An Jun, 1985:41-7). The Manchus have had a very interesting and unusual history. They grew out of tribes of hunters and warriors in the late sixteenth century. United under the leadership of Nurhaci (Qing Taizu; lived 1559-1626), the founder of the “Later Jin” khanate (1616) and the father of the first emperor of the Qing empire, Hong Taiji (Qing Taizong; lived 1592-1643), they emerged as a political force at the very beginning of the seventeenth century2. It was exactly in that 2 In fact, the real name of the creator of the Qing empire is unknown. The name, under which the first Qing emperor is known, is rather a title than a name, and it literally means “Prince Hong” (Crossley, 1997:208).
gorel-1.p65
7
4/17/02, 8:56 AM
8
part one
period that the Manchus began their first invasions. They finally entered China in 1644 and seized Peking (Beijing) which was the seat of the Ming dynasty. In the year of conquest of China the Manchu population num bered 300,000 while the Chinese population ran to 300 million (Sidikhmenov, 1985:9). The number of the Manchu forces directly involved in the con quest is estimated as approximately 120,000-150,000 of skilled, armed individuals opposing the army of the immense Chinese empire. In fact, sources give different figures for the fighting force of the Manchus, and it is very difficult to be precise in estimating the number of Manchu military troops (Crossley, 1990:13, 232). After the occupation of Peking, the capital of the Ming empire, the Manchu army reached Central and South China, installing mili tary garrisons in the conquered territories. After that the Manchus advanced to Northwest and Southwest China until they occupied Xinjiang (1755), which became the limit of the expansion of the Qing empire (Crossley, 1990:13). It was accepted until recently that the Manchus, although rulers of China for more than 250 years (1644-1911), were subsequently absorbed by China’s superiority, both cultural and numerical. 3.1. The Jurchens and their Ancient History By the mid sixteenth century, which was the time of Nurhaci’s coming into the world, numerous Tungusic tribes, known as the Jurchens, were settled in the vast spaces of Northeast Asia, to the east from the Great Khingan Range to the valleys of the Liao, Songari and Tumen-ula rivers and their tributaries. The Jurchens—like their ancient predecessors long before them—occupied the mountain regions and river valleys of Manchuria (which in present times en compasses the modern provinces of Liaoning, Jilin and Heilongjiang and which is now called the “northeast”, Dongbei, by the Chinese), as well as the Russian “Maritime Province” (Primorskii Krai) (Larichev, 1998:34). Those territories must be enlarged by the lands of Northern Korea as far south as Hamhu�ng (which were not in cluded in the Korean state until the end of the 14th century) (Crossley, 1997:14). The Jurchens had complete dominion of the territories that were located along the Songari, Ussuri, Nonni, Yalu, and Amur rivers,
gorel-1.p65
8
4/17/02, 8:56 AM
9
prolegomena to manchu studies
as well as of those, located within the boundaries of the mountain region of Changbaishan and the extensions of the Sichote Alin Range (Larichev, 1998:34). Their lands were famous for ginseng, furs, honey, and wax—white and yellow, and also, for pearls which were highly valued because of their size, rare clarity of colour, and roundness. Jurchens were engaged in farming and cattlebreeding, hunting, fishing, and gath ering. They were not united under the government of a single ruler. In former times they had a powerful state, the Jin empire (in Jurchen, “Anchun;” in Manchu, “Aisin”—the “Gold,” “Golden”), which in cluded Northern China (the centre of the Jin was located near the modern city of Harbin). That state lasted from 1115 till 1234 and fell under attacks by the Mongols3. The history of the Jin empire was described in “Jinshi”, the dy nastic chronicle of the Jurchens. This dynastic chronicle was com piled and first presented to the last Mongolian emperor of the Yuan dynasty, Toghon Temur (in Mongolian), in the late autumn of 1344. Written in Chinese, in accordance with the cannons of composition of dynastic chronicles, “Jinshi” consisted of 135 chapters. It contained, in particular, “The Main Annals”, “Benji,” as well as special chap ters devoted to geography, rites, army, economics, and the admin istrative apparatus of the Jin empire. The chronicle also included 73 chapters where biographies of outstanding figures of the empire were given. The translation of this chronicle into Manchu was undertaken later. Its completion under the title “The History of the Jin Dynasty, Ruled in northern China since 1114 till 1233” can be dated as early as the end of the first part of the 17th century. In the first part of the 19th century, this chronicle was translated into Russian by the Russian manchurologist G.M. Rozov (lived 1808-53). Recently this Russian version of the “Jinshi” has been published under the title “Istorija Zolotoj Imperii” [A History of the Gold (Golden) Empire], with historical comments written by A.G. Malyavkin (Malyavkin, 1998:233-84). The publication has been edited by V.Ye. Larichev who also wrote a brief survey of the early history of the Jurchens 3 P. Crossley gives another temporal limit of the existence of the Jin empire, i.e. from 1121 to 1234 (Crossley, 1997:208). According to all available data, the Jurchen leader Agåda came to power in 1114. In 1115 he declared himself the emperor, and created a new empire, which he called the Jin (Gold or Golden) dynasty (Kessler, 1993:13, 130; Larichev, 1998:84-5; Malyavkin, 1998:235).
gorel-1.p65
9
4/17/02, 8:56 AM
10
part one
that preceded the formation of the Jin empire. This sketch is fol lowed by the bibliography, which includes a number of articles and books previously, I assume, unknown to Western scholars (Larichev, 1998:34-87). A.G. Malyavkin undertook an investigation of the two versions of the chronicle, Chinese and Manchu. He came to the conclusion that the Manchu version was an original composition, aimed to praise deeds of the Jurchens’ forefathers and to affirm the greatness of the Jurchens. Structurally, the Manchu version consists of extracts from “Benji” and the section of “Bibliographies,” including scant infor mation from some other sources (Malyavkin, 1977:96-100). The origins of the founders of the Jin empire, the Jurchens (or earlier the Nüzhens), is a matter of debate to a certain degree. Look ing back to the Early Middle Ages, one can notice that the name “Jurchen” appeared in Chinese records, dated as early as the first half of the seventh century, in Tang times (618 to 907 AD) (Larichev, 1998:34). According to other scholars who rely on Song-era records, the name did not appear until the Later Tang dynasty (923 to 936 AD) (Kessler, 1993:129). In Chinese dynastic chronicles, which, however, could not be considered in many cases as directly helpful, the origins of the Jurchens was traditionally connected with the ancient Proto-Tungusic tribal groups of “Sushen,” “Yilou,” “Wuji,” and “Mohe,” who from time immemorial, replacing one another, inhabited the vast territo ries of Manchuria and the Russian Far East (Crossley, 1997:215; Larichev, 1998:34). Most historians believe that the Mohes were the nearest descen dants of the Jurchens. According to another point of view (see Sun Jinyi, et al, 1987), the Nüzhen (and later the Jurchen) were clearly descended from the Sushen, the oldest tribal group which is men tioned in pre-Qin texts as existing during the epoch of the legend ary Shun emperor, in the late third millenium BC (Kessler, 1993:129)4. According to V.Ye. Larichev, who currently devotes much atten 4 In fact, no accurate dates exist for the rule of the legendary Shun emperor. Chinese historiography suggests that the Sage King Shun ruled from 2255 to 2205 B.C. The King Yao (2356 to 2255 B.C.) was another legendary ruler of ancient China. They are both constantly mentioned in ancient books as exemplary rulers. Together with the King Yu, they had formed the “trio” of the rulers of the socalled “Golden Age” of China (Malyavkin, 1998:258).
gorel-1.p65
10
4/17/02, 8:56 AM
11
prolegomena to manchu studies
tion to studying ancient cultures, established in Northern (Northeast ern) China at different times, peoples in the East of Asia knew about “Sushen” (the names “Sishen”, “Jishen” also appear, albeit rarely) long ago. The name “Sushen” was mentioned for the first time in records, devoted to remarkable events, that happened in the times of the Sage king Shun (“Shiji,” jian “Uda benji;” “Huainanji,” jian “Yuandaoshun;” “Didai liji,” jian “Shaojian”). It is briefly mentioned that in 2021 BC, the embassy of Sushen had arrived to the court and presented arrows with shafts, made from the “ku” tree, and with coarse arrowheads, made from the “nu” stone. The date itself is very questionable. It is rather difficult to establish with certainty that events, such as the one happening in 2021 BC, actually occurred at a fixed point in time during the late Stone Age (1998:35). The Sushens were often mentioned in Chinese records of the Zhou period (Larichev, 1998:36-7)5. According to the opinion of some orientalists, there is no direct evidence to support this information (Crossley, 1997:215). Thus, the Japanese scholar H. Ikeuchi suggested that embassies of the Sushens never arrived to imperial courts in the pre-Han period, and all in formation about them in the Zhou period should be considered as fabricated (Ikeuchi, 1930). In contrast with that, V.Ye. Larichev believes that the Sushens played a very important role in the East of Asia not only during the Zhou period, but even before, during the Shang-Yin period (1766 to 1027 BC) (Larichev, 1998:36-7)6. There is no evidence in Chinese records that the Sushens arrived to the court during the Qin and the Han periods (221 to 207 BC and 206 BC to 220 AD). Obviously, they preserved sovereignty from the Han empire. During the Wei dynasty (220 to 264 AD) the number of times the arrival of Sushen ambassadors to the court is mentioned not only 5 Traditionally, there have been differing historical opinions concerning the temporal limits of the Zhou period. According to A. Kessler, the Zhou era lasted from 1100 to 881 B.C. or to 221 B.C. if one includes “Spring and Autumn” and “Warring states” in that period (Kessler, 1993:14). P. Crossley dates this period as existing from 1050 to 221 B.C. (Crossley, 1997:215). According to A.G. Malyavkin, the traditional point of view is that the Zhou period lasted from 1122 to 249 B.C., but he also mentioned this period as existing from 1027 to 247 B.C. (Malyavkin, 1998:250, 255, 270). 6 A. Kessler dates the Shang period as existing from 1650 to 1100 B.C. (Kessler, 1993:14).
gorel-1.p65
11
4/17/02, 8:56 AM
12
part one
increased, but invaluable and detailed ethnographical descriptions of the people and lands they possessed also appeared. Historically, the fact of the arrival of the second embassy of the Sushens to the Wei court (262 AD) is a very important one (the first embassy dated from 236 AD). When arranging this event, the Chinese did not find the tribe of Sushen in South Manchuria. In reality, in the lands where the Sushens were considered to be settled, they found a tribe the name of which sounded, according to Chinese pronunciation, as “Yilou.” Hence a very important question arises. Did the “Yilou” of the Wei period coincide with the Sushen of the Zhou period? Bearing in mind, that every time the Chinese emperors wanted to eulogize their own rule they mentioned the arrival of the Sushens to the court, in this particular case, the Yilou might be identified with the Sushen deliberately, in order to enhance the prestige of Sima Zhao who was the real ruler of the last years of the Wei dynasty (the last emperor, Cheng Liuwang, did not possess real power) (Larichev, 1998:38-40). Beginning from the fourth century, the name “Sushen” completely disappeared from Chinese records. The tribes which inhabited the largest portion of Manchuria and what is known in the present-day as the Russian Far East, came to be called “Wuji” during the North ern Wei period (386 to 534 AD) and “Mohe” during the Sui period (586 to 617 AD)7. According to some scholars, “Wuji” is the earliest name for a people who can be confidently identified as predecessors of the Jurchens (Crossley, 1997:215). Traditionally, the Mohe are consid ered as direct descendants of the Wuji. By the fifth century AD, information about the Mohe people in creased to such a degree that it became possible to speak about a whole group of related tribes, seven in number. Each tribe occu pied a strictly appointed territory. In the extreme south of territories where the Mohes settled, the tribe Limo (Sumo or Songmo) lived. They possessed the lands ad joining the river basin of the Upper Songari, north Jilin, as well as Baishan, the mountain-mass, located southeast, close to Mudangjiang. The Songmo were the most powerful of the Southern Mohe group. 7 P. Crossley gives the temporal limits of the Northern Wei period as existing from 465 to 535 A.D. (Crossley, 1997:15).
gorel-1.p65
12
4/17/02, 8:56 AM
13
prolegomena to manchu studies
The Bodo (or Gudo) were settled in the place of confluence of the Nonni and Songari rivers, north of the Limo Mohe, in the region of the modern city of Bodune. Further to the north-east, near the modern city of Harbin, along the southern bank of the Songari river where one of the most known tributaries of Songari, the Alchuk river runs, the lands of the tribe Anchegu Mohe were located. To the east of the Gudo Mohe, on the banks of the lower and middle reaches of the Songari and Hurha rivers, the tribe Fune Mohe lived. The tribe Haoshi (Gushi) Mohe occupied the area, located near the lower reaches of the Songari river, in the region of the modern city of Sansing. The most powerful of the Northern Mohe groups were Heishui (the name of the Amur river in Chinese; Heisui in Manchu), who were settled along the bank of the Amur river, in the territo ries located between the mouths of Songari and Ussuri. The Mohes’ way of life, as it was described in Chinese records, strongly resembled that of the Sushens. According to the dynastic chronicles of theYuanwei and Tang eras, the Wuji and Mohe did not know how to build houses above ground, instead they built dwellings of semi-subterranean style, a kind of earth-houses. Those dwellings were round in shape, and from above they looked like sepulchral mounds. Such a dwelling had a central chimney hole through which people normally entered by a ladder with several steps. Rich houses had ladders with nine steps. Without a question, such houses were dark and uncomfortable, but they retained warmth in the harsh climate of the region (Crossley, 1997:19; Larichev, 1998:53). In Chinese records it was noted that the Mohes were the dirtiest people in comparison with the other eastern peoples. In particular, aliens were greatly astonished that the Mohes washed their hands and faces with their own urine (Larichev, 1998:53). A number of houses formed a settlement, which was surrounded for protection by defensive walls, made like a dike, similar to a ram part. In summer the Mohes left their winter houses to search for new places, rich in water and grass. A semi-nomadic way of life they lead in summer, forced them to build special dwellings similar to light portable tents. By winter, they returned to their fortified settlements. The Mohe tribes can be characterized as having mostly a seden tary way of life. Pig breeding was the main branch of their stock
gorel-1.p65
13
4/17/02, 8:56 AM
14
part one
raising (the same can be said about the Sushen). The richest of them possessed large flocks of pigs, which mounted to hundreds of heads. They liked eating pork, and pig-skins were used by them for sewing winter coats (again like the Sushen). The Mohes also bred dogs, from whose skins they sewed fur coats. At the same time, they bred horses. The Mohe warriors could not imagine their life without a horse, their foremost ally. According to their funerary practice, they buried horses with the interred. The Mohe were mainly engaged in developed agriculture, as they used a plough, pushing it in front, and pairs of horses as drought power. They sowed wheat, millet, rice plants, and soybean, and used rice for preparing special liqueur. Being strong and sturdily-build, they never got drunk. The Mohe were also engaged in hunting, fishing, and gathering. Women were widely engaged in weaving. Under warm clothing, they preferred wearing dress made from linen. The Amur Mohe wore headdresses—probably, a sort of full uniform—decorated with pheasant feathers and wild boar fangs. In the South, the Songari Mohe horrified their enemies by their hats, decorated with tiger and leopard tails. When burying corpses, they did not make coffins (in contrast to the Sushens). Moreover, they digged graves only in spring and sum mer. They lowerred a corpse to the bottom of a grave, put funeral meals in front and performed offerings. In winter, corpses were left in places distant from a settlement, in open air, and funeral rites were then carried out (Larichev, 1998:53). The Mohe warriors were very famous in northern Asia for their courage and military skills, the main ones of which were horseman ship and archery. They were superlative archers, who were trained to hunt from horseback, like other northeasteners, from their child hood. The Mohe warrior’s primary weapon was a bow. Their ar rows never missed their aims, horrifying neighbouring peoples, which the Mohes constantly raided. They knew no fear. What is interest ing to mention is that some of the Tang’s Army most important military commanders were from the Mohes tribal chieftains, who helped the Chinese gain a victory over the Liaodong8. 8 The area, located east from the Liaohe river, was called Liaodong prior to the 17th century. In Chinese records, this was the name for southern Manchuria from the times of the Qin dynasty (from 246 to 207 B.C. according to A.G. Malyavkin, and from 221 to 207 B.C. according to A. Kessler) (Malyavkin,
gorel-1.p65
14
4/17/02, 8:56 AM
15
prolegomena to manchu studies
The length of a bow—in comparison with that of the Sushens— decreased in size and did not exceed one metre, and the length of an arrow was about 40 cm. However, those changes did not weaken the killing power of the Mohe warrior’s weapon. The point is that the Mohe warriors decorated their bows with horny plates, which in reality were not for decoration, but for increasing the resiliency and therefore the power of bows. The Mohes, like the Sushens before them, used extremely strong poison to smear their headarrows (Larichev, 1998:53-4). The information concerning the social life of the Mohe is very scant. Definitely, there were poor and wealthy families among them. Separate tribes continued keeping their independence one from an other. They fought with each other for territories, livestock, and property, and abductions were very frequent. The Mohes called their tribal leader “damofo mandu” (chin. da “great”), as one can see further, the Southern Shiwei, who can be identified as people of Tungusic descent, called their tribal chieftains “yumofo mandu” (see section 4.1). Tribal chieftains had complete power over their fellow-tribesmen and in addition possessed a considerable number of slaves. Slavery, as a social institution, was probably of patriarchal character (Larichev, 1998:54). The language spoken by the Mohe was Tungus-Manchu. What is important to mention is that the language of the Sushen could also be referred to as proto-Tungusic. During the Tang era, the Mohe, similar to other peoples of north eastern Asia, were subjected to constant political and military pres sure from Tang rulers. Soon after the Koguryo� state of Korea had been defeated by the Tang empire (668 AD), a large portion of the Koguryo� people fled into the lands of the Sumo Mohe9. Soon a lot of towns, surrounded by defensive walls, arose there. Around 700, a new state, “Parhae” (chin. Bohai), raised from the ruins of Koguryo�, was established. It was the leader of Sumo Mohe, Cicik Zhungxiang, who was considered the creator of Bohai. He found a way to alle viate the pressure of Tang, which was compelled to recognize his lineage as a local hegemon and to permit the establishing of a sort 1998:260; Kessler, 1993:14). It is now the modern province of Liaoning in China. 9 The early-feudal Koguryo� state, which included a large portion of lands, neighboring south Manchuria, existed in those times. In 668 A.D. the Tang commander Li Ji, in alliance with the Silla state of Korea, destroyed Koguryo�.
gorel-1.p65
15
4/17/02, 8:56 AM
16
part one
of a demi-state. Later, his grandson, Uazhi Da Tuyu, declared himself the emperor of Bohai, which in the course of time became highly cultured and enlightened, and widely known beyond the borders of the country. The Parhae (Bohai) state—a deserving successor of the culture and power of Koguryo� and the tribal league of the Songari Mohe—flourished for 228 years until it was destroyed by the Qitans (926 AD) (Shavkunov, 1968; Crossley, 1997:18; Larichev, 1998:534). As for the most Northern of the Mohe tribes, the Heishui (Amur) Mohe, its fate was totally different. The Heishui Mohes created their Army, built fortifications along their northern and southern bounderies and successufully repulsed the attackes of the Bohai to invade their territories. The fight of the Bohai against the Northern tribes lasted for several decades, but the Heishui Mohe had preserved their independence to the end. During the tenth century, the Heishui Mohe had drammatically increased their military power and political influence upon the peoples of northern Asia. By that time, they controled not only the territories of the basin of the Middle Amur, but also the upper reaches of the Songari river and probably the valleys of the Ussuri and Suifung rivers. It was during this time that the tribe of Nuizhen became prominent among all the others by their power. The name “Jurchen” appeared for the first time and all Heishui Mohe came to be called by the name “Nüzhen.” When the Liao empire of the Qitans destroyed Bohai in 926, it was the Jurchens who started strug gling for independence of the peoples of the Far East against the nomads of central Asiatic steppes. At the same time, the Jurchen tribes struggled fiercely among themselves in order to seize power and unite all groups until one of them, the Wanyan, achieved supremacy. The leader of this group, Agåda, played a great role in the for mation of the Jurchen tribal league. He constantly pursued a policy of independence from other tribes of Manchuria, as well as from the Liao and Koryo� states. Agåda came to power in 1114, and shortly began rebelling against the Liao empire, which suffered one defeat after another. In 1115 Agåda formally declared himself the emperor of a new state which he called the Jin (Gold or Golden) empire. By the end of Agåda’s rule (he ruled from 1115 to 1124), the Jurchen forces had captured most of the Liao territories, including their southern capital (present-day Beijing) (Vorobiev, 1975; Kessler,
gorel-1.p65
16
4/17/02, 8:56 AM
17
prolegomena to manchu studies
1993:129-30; Crossley, 1997:19-20; Larichev, 1998:55-86)10. The Jin empire was on very fragile terms with the Northern Song dynasty (960 to 1127 AD). Their hostilities were came to an end in 1127 when the Jin forces conquered the Northern Song capital of Bianjing (present-day Kaifeng in eastern Henan province). The re mains of the Song government shifted south and established the Southern Song dynasty (1127 to 1279 AD). The Song empire lost their territories in the Hebei, Henan, Shanxi, and Shaanxi provinces according to the peace agreement the two empires reached in 1141. However, their hostilities continued until 1165, when the two em pires made a peace (Kessler, 1993:130). During the Jin dynasty’s rule, the Jurchens, including the Mongols who were under the suzerainty of the Jurchens for 100 years, sub jected many peoples of northern Asia. When around the end of the twelfth century the Jin empire began to suffer from internal disor ders and problems, the Mongols, aware of the Jurchen’s weaknesses, began their campaign against the Jin dynasty. By 1214, the Mongols laid siege to the Jin capital of Zhongdu (present-day Beijing). The Jurchens agreed to pay tribute, and at that time they made peace with the Mongols. In 1234, Ögödei, the third son of Genghis Khan, completed the conquest of the Jin empire (Kessler, 1993:148,154). After the Jin empire, Jurchen tribes were widely dispersed through vast territories of Manchuria and northern China (Crossley, 1997:208). By the mid sixteenth the Jurchens divided into numerous tribal federations (in Manchu, “aiman”). Several tribal federations consti tuted more large clan-tribal unions or confederations (in Manchu, “culgan”)11. Both in tribal federations and confederations there was no peace. Permanent discords and internecine wars destroyed them. By the time Nurhaci began his campaigns “to unite various peoples under one rule”12, three large antagonistic unions existed: Manchu, Golmin-Shanyan-alin, and Hålun. They were in a state of perma nent war with one another. The Manchu confederation included the Jianzhou tribal federation (aiman) of which Nurhaci’s grandfather and father were clan headmen, and into which Nurhaci himself was born. 10 Subsequently the history of the Wanyan was interpreted by Jin rulers as the history of the whole Jin empire. 11 J. Norman translates the word aiman as “a tribe” and the word culgan as “an assembly,” “a league” [NL:9, 48]. 12 Quoted in Pamela Kyle Crossley, “Orphan Warriors,” 1990.
gorel-1.p65
17
4/17/02, 8:56 AM
18
part one
The Hålun confederation which was influenced heavily by the Mongols (some Hålun tribal headmen had the Mongols among their ancestors), consisted of the Ula, Hada, Hoifa, and Yehe tribal fed erations, and all of them clashed with the Jianzhou Jurcherns from time to time. The Golmin-Shanyan-alin union included the Yaluula, Neyen, and Jusheri tribal federations, and they were defeated by Nurhaci and added to his possessions in the first expansion of his influence (Kuznetsov, 1985:3-5). The Jurchens bordered Mongol lands, China, where the Ming dynasty ruled (1368-1644) and the Korean realm. To subjugate the Jurchens, governors of the Ming Empire used the practice of distributing official ranks among Jurchen foremen. The Ming authorities announced a clan headman who was rewarded with a certain official rank as the head of a guard or a post. Corre spondingly, the Ming administration considered a Jurchen commu nity whose leader was given a certain official rank and a seal as their vassal. Thus, a place where Nurhaci and his fellow-tribesmen lived, the Chinese called “Jianzhou post,” and Nurhaci’s forefathers had the rank of the “head of Jianzhou post.” Those who possessed a certain official rank could go to the Ming court and trade within Chinese fortresses. The Jurchen foremen brought gifts to the Ming court, however, those gifts, according to traditional Chinese diplo matic protocol, were classified as tribute even though Chinese gifts were given in return. Nevertheless there was no stable peace on the boundary between China and the Jurchen lands. In 1583 Ming military troops made punitive expeditions within the Jurchen bor ders. As a result of those offensives, Nurhaci’s grandfather Giocangga and father Taksi were killed (Kuznetsov, 1985:3-5). 3.2. The Manchus’ Ethnicity As shown in recent studies, in the early period of Nurhaci’s reign the Jurchens were not an ethnic conglomeration. They included not only various Jurchen tribes whose ancestors had created the Jin empire in Northeast and Northern China, but also acculturated descendants of Mongolian, Chinese and Korean peoples and diverse Northeastern tribes like the Evenks, Solons, Orochons, Golds, and Sibes. These tribes, as the Tungusic class of the Altaic language group, were genetically and linguistically close to the Jurchens proper, and were slightly influenced by the Chinese, Mongols, and Kore
gorel-1.p65
18
4/17/02, 8:56 AM
19
prolegomena to manchu studies
ans. Hong Taiji (Qing Taizong), the second khan of the Later Jin (1627-35) and the first emperor of the Qing dynasty (1636-43), claimed in 1635 that the majority of the banner population, mainly Jurchens, were to be called Manchus. The new identity, in this way, also included several components and was not a homogenous eth nic unit (Crossley, 1990:16). Thus, the Manchus, were drawn from a culturally diverse collection of peoples, and the term “Manchu” itself had more political sense than ethnic-anthropological meaning (Stary, 1994:470). 3.3. Clans in the History of the Manchus When Hong Taiji declared himself the emperor and created the Qing empire (1636), various groups of population differing from the point of view of their ethnic affiliation, economic order, language and culture, were united under the empire. Nevertheless there were some basic similarities in their social organization and cultural life, such as clans, oral history and shamanism, their ancestral religion (Crossley, 1990:34). A clan (in Manchu, “mukån”) was the principal entity in the so cial organization of the Manchus and of the Jurchens before them13. The role of a clan had been changed in the course of time. Before and during the Manchus’ conquest of China, clans governed the everyday life of the people. Clan headmen were very often shamans themselves, in order to show the best way of farming, hunting, gath ering and how to make war on antagonistic tribes. After the con quest, when aggressive wars ceased and the polity revealed a ten dency to transform from a “booty” into a bureaucratic state, economic conditions varied sharply and the role of the clans was diminished. They began to hamper the economic and social devel opment of society until they finally changed into political and cul 13 Several definitions of a Manchu clan exist. According to S.M. Shirokogoroff, a clan, as a social entity, is characterized by some principal features. All members of a clan are descended from a single male ancestor, which results in a necessary blood relationship among them. In addition, a clan has common clan spirits and taboos, the most important of which is exogamy, and furthermore, it is obligatory to have a name for each clan (Shirokogoroff, 1973:16, 31). P. Crossley maintains that not all clans could trace their origin to a single ancestor and even in the late sixteenth century biological kinship was not an obligatory characteristic for absolutely all clans, but she manifests solidarity with Shirokogoroff in acknowledging that the clan could not exist without a name and “the clan name (hala), which was most often a toponim of ancient or recent vintage, was the key to the origin of all mukån sharing that name” (Crossley, 1990:36).
gorel-1.p65
19
4/17/02, 8:56 AM
20
part one
tural symbols, with affiliation becoming a point of pride by the nine teenth century (Crossley, 1990:34). The history of the Jurchens and later of the Manchus was closely related to the clans, or to be more precise, to their separate histo ries. The history of the whole empire was interpreted by rulers as the history of the imperial clan, the Aisin Gioro (the Golden Gioro), which was created by Nurhaci after the middle of the seventeenth century. Nurhaci’s clan, Gioro, pretending to be a clan of ancient history, had many sub-branches throughout the Northeast. It was very important for Nurhaci to invent his own clan in order, firstly, to claim his clan’s superiority over others, and, secondly, to show that his clan was the most ancient, tracing its origin to the Jurchen Jin dynasty. Bearing in mind that all clans had murky, poorly docu mented histories at least before Nurhaci’s time, it is clear why Nurhaci and his descendants exerted every effort to invent a clan history for themselves (Crossley, 1990:32-3)14. Considerable efforts were made during the reigns of Emperors Yongzheng and Qianlong (1723-1735 &1736-1795 respectively), to reestablish the social and spiritual role of the clans. The rulers understood the necessity of well-documented histories of the clans and therefore they arranged publication of the genealogies of the ancient clans, especially of those that had formed the basis of the Eight Banners, “the socio-military organizational foundation of the Qing state” (Crossley, 1997:207). The most important of those histories, such as “General History of the Eight Banners” ( published in 1736), “Comprehensive Genealogies of the Clans and Lineages of the EightBanner Manchus” (published in 1745), “Researches on Manchu Origins” (published in 1783), came out in the reign period of the Qianlong emperor (Crossley, 1990:21). In the Jurchen’s times the ancestral religion, shamanism, was one of the prominent elements in their social organization and spiritual life. Often being clan headmen, shamans sanctified economic activi 14 As evidenced by some historical records, Nurhaci’s own lineage could be traced to Möngke Temür, the powerful leader of the Jianzhou federation, in whose name one can ditect an obvious Mongolian influence. In accordance with folk tradition, Möngke Temür’s father or grandfather had been a leader of the Ilantumen Jurchens (Ilantumen being a place at the Songari-Amur confluence) before their migration to the region of modern Hunqun in the South at the end of the fourteenth century. There is historical evidence that Möngke Temür was killed in the battle with other Jurchen tribes on the North Korean boundary in 1433 (Crossley, 1990:323).
gorel-1.p65
20
4/17/02, 8:56 AM
21
prolegomena to manchu studies
ties of all members of a clan, helping them communicate with their spirits. With the reorganization of the polity into a bureaucratic state, when the population had no further need to support themselves by traditional activities such as hunting, fishing, farming, gathering, and fighting against antagonistic tribes and peoples, the Manchus needed to communicate less often with their helping spirits, and the role of religion, as well as that of a clan, was also sharply diminished. Despite considerable efforts that were made to re-establish shamanistic rites, especially in the time of Emperor Qianlong, the language of those rites had become absolutely incomprehensible for young bannermen. 3.4. Banners and Garrisons as Major Government Institutions of the Manchus Banners and garrisons were major government institutions both in the Later Jin khanate and the Qing state. A banner (in Manchu, “gåsa”) was originally a military organization of the Jurchens and later Manchus, created by Nurhaci to unite disparate groups of popula tion within the new identity. Those military units were called “ban ners” because each of them carried a distinctive flag. Later a ban ner became an administrative and social organization as well. At the very beginning of the sixteenth century Nurhaci created four ban ners (yellow, white, blue, and red). The majority of specialists agree that the time of their creation can be dated from 1601. Subsequently, with the increase in population, he founded four more military units. In 1616 each banner was split into two, one being “plain” and the other being “bodered” (Crossley, 1997:207). The whole socio-military organization came to be called the Eight Banners (in Manchu, “jakån gåsa”). Every banner was distinguished by its colour, so that the whole army included the Plain Yellow, Bordered Yellow, Plain White, Bordered White, Plain Red, Bordered Red, Plain Blue, and Bordered Blue banners. Later, eight new banners were created for the Mongols and as many again for the Chinese-martial (in 1642)15. Ultimately the regular forces numbered twenty-four banners. 15 P. Crossley referred to the Chinese-martial as those who were in most cases of indeterminable Jurchen, Chinese or Korean descent and were located within the Ming pale of Liaoning peninsula, primarily modern Liaoning province. If military units of the Manchus proper were based upon “companies,” originally clan organizations, by contrast, a Chinese-martial division of the Eight Banners was based upon created “companies” (Crossley, 1990:6; 1997:29, 203-5).
gorel-1.p65
21
4/17/02, 8:56 AM
22
part one
Early banners created by Nurhaci developed from companies (in Manchu, “niru”) which borrowed their organizational principles from the clan hunting group. A clan was the ideological and structural base of the Manchu Army: a company and a banner. Clan head men were often company leaders called by the ancient term “nirui ejen” (“lord of the arrows”). A process of reorganization that started in Hong Taiji’s reign and continued into Shunzhi’s and Kangxi’s times, caused many innovations in the structure of companies whose leaders were called by a new term—“nirui janggin” (“captain of the company”) (Crossley, 1990:36). According to different sources, the number of soldiers that constituted one niru, varies radically, from a hundred to three hundred (Crossley, 1990:231-2)16. In Hong Taiji’s times (1627-1643) and then during Shunzhi’s (1644-1661), Kangxi’s (1662-1722) and especially Yongzheng’s (1723 1735) reigns, as a result of diminution of the role of the clans in the structure of banners, companies began to loosen their direct asso ciations with clans, and “clan headmen were no longer recognized by the court as banner officials, but were chosen according to the ostensibly meritocratic standards that were permeating the banners as a professional institution” (Crossley, 1990:37). A banner, as an institution, was borrowed from Turko-Mongolian hereditary military servitude. According to its regulations, a soldier was a slave of his owner (Sinor, 1981:133; Crossley, 1990:15). At the very beginning of the seventeenth century when the Manchu army, the Eight Banners, was organized, a bannerman denoted a man who was enrolled in one of the banners. A bannerman should possess certain military skills, among which the most valued ones were horsemanship and archery. Like all Inner Asian conquer ors, a bannerman was able to shoot from horseback. The personal relationship which existed between a Manchu soldier and his owner was a continuation and development of the Turko-Mongolian sla very when a soldier was never free (Crosley, 1990:15). Even by the end of the nineteenth century, “bannermen,”—as the Chinese writer Lao She, originally a Manchu of the Sumuru clan, wrote in his 16 J. Norman translates the word niru as “a banner company of a hundred men” (NL:216). According to V.S. Kuznetzov, “niru,” on average, had three hundred soldiers. He also maintains that five companies (five “niru”) formed one regiment (in Manchu, “cale”), five “cale” constituted one corps (in Manchu, “gåsa”), and a banner consisted of two “gåsa”. Bearing in mind, that niru had three hundred soldiers, a banner should number about fifteen thousand soldiers (Kuznetsov, 1985:3-5).
gorel-1.p65
22
4/17/02, 8:56 AM
23
prolegomena to manchu studies
famous novel “Hongqi zhi xia” (“Beneath the Red Banner”),—“had no real freedom; they could not quit their original Banner or travel outside of Beijing;” “a bannerman was expected to join the Impe rial Army, ride a horse, shoot with a bow and arrow, and defend the Qing dynasty” (Lao She, 1982:53-4). In the early seventeenth century ethnic affiliation was not a prin cipal qualification for a man to be considered as a bannerman: “a bannerman might be a Manchu, a Mongol, a Northern tribesman, a Korean or Chinese transfrontiersman, who had joined Nurhaci’s band either to avoid annihilation or to profit by the expansion of the Later Jin khanate (1616-1626)” (Crossley, 1990:14). Affiliation with the nobility was not obligatory for a bannerman. From the very beginning of the presence of such a military insti tution as a banner, each Manchu bannerman was well compensated after the conquest, first with booty and then with rice and cash. The bannermen, as a group, enjoyed some privileges until the middle of the nineteenth century which the rest of the population of the Empire didn’t have. Even at the end of the nineteenth century, when the Qing em pire was rotten to the core and the Imperial government could hardly provide bannermen with allowances, each family of a poor Manchu soldier received a fixed income of three ounces of silver and a cer tain portion of rice. Lao She, who was born in one such Manchu family of a poor soldier, wrote that this fixed income was a semblance of a salary, and because the Manchus were not allowed to have another job apart from being a soldier in the Imperial Army, all poor Manchus “felt that there was a noose around our necks which was getting tighter all the time” (Lao She, 1982:32). By the middle of the eighteenth century, when the Manchus fin ished their aggressive wars against antagonistic tribes and peoples and pacification of Xinjiang in 1755 became the last expansion of the Empire, the Eight Banners had lost their military role in the state. New economic, social and cultural situations came into existence in the Qing empire. The state was no longer booty-based. A bannerman population was increasing constantly, whenever the number of vacant posts in the Imperial Army was limited. There arose staff competition for banner company appointments and hereditary cap taincies. In accord with Qing military regulations, only two older sons of a Manchu family could be taken into the military service. The others were forced to remain unemployed. If a family had one
gorel-1.p65
23
4/17/02, 8:56 AM
24
part one
or two unemployed persons it made life extremely hard. Bearing in mind that a man might wait for a vacant post for a long time, it is understandable that many young men, being life-long unemployed, sensed a loss of freedom and self-respect. By the end of the ninteenth century Manchu society had divided sharply into two opposing camps: Manchu nobility whose ancestors accumulated great wealth enjoyed their lives, while poor Manchu soldiers could not make ends meet. Lao She wrote about the life of the Manchu society at the end of the nineteenth century: “We evolved a unique style of living. Those with money wanted the best of ev erything; so did those without in spite of their poverty. So all sank or floated in the stagnant pond of their pleasure-seeking lives” (Lao She, 1982:32). Nevertheless both the Manchu nobility and poor Manchu soldiers continued having certain privileges, in some cases, very questionable. A garrison was another fundamental entity in the military regimes of Later Jin and Qing empires. Hundreds of garrisons were installed throughout successively conquered territories, and their size ranged from about five hundred bannermen to about thirty thousand. Enrolled bannermen lived in garrisons with their families, support ing themselves from proceeds of their lands and the allowance which the court paid them monthly in cash (silver stipend). A garrison was a rather closed unit; after the early eighteenth century no bannermen could move away from the garrison to which he was assigned and in which he was born. Activities of the bannermen within each garrison were strictly limited. Diversity of dialects and subdialects was closely related with the multitude of garrisons. But those dialects were rarely written and changed to a large degree as a result of their contacts with the Chinese language (Crossley, 1990:6, 23, 28). 3.5. Manchu Identity The old Jin khanate and the Qing state at the beginning of the initial period of conquest were “booty” in their origin. The booty was distributed by the khan among his retainers (amban and beile), who in their turn divided it with bannermen17. Even slaves were given 17 J. Norman translates the word amban as “a high official,” “dignitary” and the word beile as “ruler,” “prince of the third rank” (NC:15, 27). According to A.G.
gorel-1.p65
24
4/17/02, 8:56 AM
25
prolegomena to manchu studies
booty for their service. The military servitude borrowed from the Turki and the Mongols, was the foundation of the banners, as the state’s institution. In that sense, the Turko-Mongolian institution of hereditary military slavery was the basis of Manchu identity in the middle seventeenth century. Clan organization of the medieval so ciety was the mechanism which governed bannermen’s military activities and controlled distribution of booty to the warriors. It should be stressed that before the reforms of Hong Taiji, the imperial clan Aisin Gioro was the owner of the banners (Crossley, 1990:15). So it would be correct to say that clans were the principal characteristic of Manchu identity. By the late eighteenth century, with the peace following the last Manchu invasions, the Eight Banners had lost their role as the main stay of imperial power (Crossley, 1990:20). Bannermen of some communities lived on proceeds from lands which they inherited from their ancestors. Members of other units had lost their lands and survived on a fixed income, the rice and silver stipend. But the latter, growing smaller and smaller in size in the course of time, was given only to those who were in the special Bannerman register. Being permanently unemployed, many of the bannermen were not enscribed in the Register and were in a state of real poverty. Thievery and desertion began to spread among the bannermen. If a bannerman made up his mind to learn to trade, he was not approved by a society in which this kind of economic activity was considered as illegal. Even at the end of the nineteenth century it was assumed that a bannerman who began to trade lost his bannerman’s sense of self-respect. Malyavkin, the word amban was used to address the governor-general who was appointed by the Manchus to govern conquered territories. This term appeared in the time of the Jurchens, but began to be used as the name for a post only in the time of the Manchus (Malyavkin, 1998:238). For some hypotheses concerning the origin of the word beile see also P. Crossley who maintains that in the time of Nurhaci this word was the title for a tribal chieftain in Manchuria. In Jurchen this word had the form of begile (Crossley, 1997: 206). Giving an explanation of the word du-bo-ji-le, which was used in “Jinshi,” A.G. Malyavkin pointed out that the Jurchen word bojile designated a leader, a chief, a head, a superior. Before the foundation of the Jin empire, the functions of bojile were not firmly established. During this period there were a lot of bojile. Agåda, as the supreme leader, was given the rank of dubojile, where du can be understood as the Solon or Manchu word da “leader,” “chief,” or it may be an inaccurate translation of the Manchu word amban into Chinese (Malyavkin, 1998:239). Obviously that the form bojile is the Chinese pronunciation of the Jurchen word begile: jur. be < chin. bo and jur. gi < chin. ji (L.G.).
gorel-1.p65
25
4/17/02, 8:56 AM
26
part one
By the end of the nineteenth century the Manchu Army was destroyed. Many officers, captains and cavalrymen did not possess riding and shooting skills, but all of them were provided by the imperial goverment with allowances. Wealthy soldiers, some of them being hunchbacks or cripples, paid others to perform their military service. Others, not possessing military skills, lived off the defence budget by hereditary right. Soldier’s widows were also provided with the Qing government’s allowances. Lao She described the situation which arose in the Empire at the end of the nineteenth century: “No wonder the French and English Allied Armies broke through the gates of Beijing and burnt down the Imperial Yuanminguan gardens. With an army composed of widows, hunchbacks and cripples living off the defence budget along with captains and cavalrymen like my eldest sister’s father-in-law and his son, how could China possibly fend off an enemy attack?” (Lao She, 1982:55). From the mid eighteenth century the Qing court, as it became fully aware of economic decline among the populations, needed to find an idea to help unite individuals within a single whole. That idea was “manjurarengge” (“Manchuness”), and two criteria were taken as a basis, military skills (riding and archery), and more importantly a mastery of the Manchu language, preferably in its spoken form18. To construct the cultural identity of the Manchu race, the idea was clearly formulated during the Qianlong reign (1736-1795), by the Emperor himself who produced special edicts and instructions to promote standard Manchu language and Manchu literacy. The Qianlong court made considerable efforts to preserve the ancestral language of the Manchus. All bannermen had to pass examinations in Manchu, and only speakers of Manchu were taken into the State apparatus. The “purity” of Manchu became a matter of the intent care of the Emperor Qianlong. The attention which Qianlong and succeeding emperors concentrated on the Manchu language “was characteristic of the shift in standards of identity from military ser vitude to cultural (and by implication spiritual) life” (Crossley, 1990:22-4). 18 The word manjurarengge “Manchuness,” the form in -ngge derived from the imperfect participle of the verb manjura- “to speak or write Manchu,” “to act like a Manchu,” “to behave in the Manchu manner” (NL:194), is written here as it is used by Crossley (Crossley, 1990:21-2). This word also occurs in the form of manjurarangge which properly complies with the law of vowel harmony in Manchu (see, for example, P.G. Möllendorff, “Manchu Grammar,” 1892:32).
gorel-1.p65
26
4/17/02, 8:56 AM
27
prolegomena to manchu studies 3.6. The Decline of the Manchu Language
Literary Manchu arose on the basis of a dialect of clans supporting Nurhaci at the end of the sixteenth century. Later this dialect be came known as the Southern Manchu dialect. The problem of creating the standard language became one of great importance, and the creation of Manchu writing was started at the very beginning of the Qing dynasty. After the conquest of China the Manchus became a minority in a huge empire where the Chinese were the majority of the popula tion. Manchu was destined to become the official language of the ancient state, but, being the language of hunters and warriors, it was lacking in bureaucratic, administrative and scientific terminology. So it was inevitable that Manchu was influenced by the more de veloped Chinese language, especially in the intellectual sphere. There was another reason for the decline of Manchu—the necessity for the two nations to communicate not only in the bureaucratic field, but also in everyday life (Stary, 1994:470). The problem of sinicization of the Manchu language became apparent even to the first Manchu rulers. Thus, it was Hong Taiji who, connecting the destruction of any dynasty with the decline of the language, urged safeguarding the “purity” of Manchu (Stary, 1994:471). Although after the conquest bannermen lived in garrisons which were closed communities, economic and cultural conditions con strained them to master at least two cultures, Manchu and Chinese. Among the Beijing bannermen, who lived in unique living condi tions under China’s cultural and numerical superiority, assimilating different Han, Mongolian, and Moslem customs, the process of sinicization and the decline of their native language was the most thorough. Russian manchurologist I.I. Zakharov showed in his Grammar of the Manchu language how heavily the phonetics of Beijing Manchus was influenced by the Northern dialect of Chinese. They articulated many of the Manchu sounds and syllables in accordance with Chinese pronunciation (Zakharov, 1879:50-61). The linguistic situation which existed among Beijing’s bannermen at the end of the nineteenth century, was accurately described by Lao She: “Ac tually, this throughly seasoned bannerman was one-half or perhaps one-third bannerman. This was not a matter of his ancestry. He knew
gorel-1.p65
27
4/17/02, 8:56 AM
28
part one
only a smattering of Manchu and always spoke Chinese, which he used for any occasional writing he did. He could neither compose classical poetry nor write formal essays or political discourses. If required to write a short song or a New Year’s couplet, he would always think in Chinese, without even considering the possibility of using Manchu” (Lao She, 1982:50). Since by the end of the seventeenth century the Manchus began to neglect their language under the poweful influence of Chinese, the ruling Qing Dynasty made considerable efforts to preserve its mother-tongue. Quite a number of monolingual and bilingual text books were prepared. A whole series of Manchu dictionaries with translation into Chinese was created. It was a paradox, but by the time Manchu rulers proclaimed mastery in Manchu as the main national idea, especially in the reigns of the Kangxi and Qianlong emperors who encouraged development of the Manchu language, the speaking of the ancestral language among the bannermen de clined (Crossley, 1990:28). According to recent studies, disappearance of Manchu proceeded in various forms on various levels of Manchu society. Within the State Apparatus the decline of the Manchu language began during the second part of the eighteenth century. Its teaching in schools was banned by a series of imperial decrees. Some new recently discov ered documents allow to conclude that the period of total loss of the Manchu language in China occurred much later than previously thought. These documents show that on the popular level, among people belonging to the lower classes, there existed a long period of bilingualism lasting until the Xinhai revolution (Stary, 1994:473-4). In accordance with the latest studies, the complete decline of the Manchu language took place after the fall of the Qing Dynasty which occurred in 1911. Disappearance of Manchu culture and language was closely related to strongly manifested anti-Manchu feelings among the Chinese population. These feelings were understandable because, for the Chinese people, the Manchus were the foreign oppressors. Consequently, everything linked with this nation, their language and culture was denigrated. The Manchus dissembled their origin and descent, especially those who traced their origin from the ruling clan Aisin Gioro. They assumed Chinese names and surnames and hid every link with the Manchu nation. In such peripheral areas of China as Manchuria and Xinjiang the
gorel-1.p65
28
4/17/02, 8:56 AM
29
prolegomena to manchu studies
Manchu language continued to be used in one form or another during the first years of the Republic (Stary, 1994:475). 3.7. Rebirth of the Manchus It seemed that the history of the Manchus, their language and cul ture was closed. But the situation has been drastically changed with the discovery of several settlements and villages in Manchuria in the 60’s. The Manchu language and culture were unexpectedly reborn. Supported by Chinese policies of the last thirty years, the Manchus have experienced a renaissance. The teaching of Manchu began in several Manchurian settlements and villages and special Manchu language courses were established at the academic level (in 1961), and later in some Universities. An impressive quantity of scientific works devoted to Manchu studies was published. In the territory of the People’s Republic of China 340 “Manchu Rural Communities” and 12 “Manchu Autonomous Counties” have been established in Liaoning, Hebei, and Jilin provinces. In several counties inscriptions on public buildings, good-luck streamers and lorries are made in two languages, Manchu and Chinese. Most material remains of Manchu culture have been carefully preserved up to now. The remnants of Nurhaci’s capitals and court residences—Fe Ala (“Old Hill”), Hetu Ala, Jiefan, Dergi Hecen (“The Eastern Capital”) have been partially restored. As it is known, there was a residence of Nurhaci and his eighth son, the first Qing empe ror, in Mukden (modern Shenyang). Now it is the “Imperial Palace Museum” which serves as an outstanding evidence that the city was the seat of the Manchu empire. The Imperial Palace preserved some elements of autochthonous Manchu culture, the inscriptions in Manchu over the doors of the various buildings in the “Forbidden City” and carvings and sculptures of dogs which were sacred ani mals in the Manchu world (Stary, 1995:16). The Imperial tombs are especially revered by the Manchus. The best known of them are Nurhaci’s tomb (Dongling “Eastern Tomb”) and Hong Taiji’s tomb (Beiling “Northern Tomb”). Yongling “Eter nal Tombs” where Nurhaci’s father Taksi, grandfather Giocangga and his great grandfather Fuman were buried, are rather less known. They are situated to the North of Hetu Ala in the Manchu Autono mous County of Xinbin, south-east of Fushun (Stary, 1995:41). Some Manchu inscriptions still exist on commemorative steles,
gorel-1.p65
29
4/17/02, 8:56 AM
30
part one
tombstones, on city doors and on mountain rock-faces, but many of them are poorly preserved or even almost eroded. Even the role of the Manchus in the history of China was revised. According to the new official opinion, the Manchu people made a valuable contribution to the defence of China’s frontier from for eign aggresion, to the foundation of a unified Chinese state, and, in general, to Chinese culture. In 1986 a monument in honour of Nurhaci, with the inscription describing him as “an excellent poli tician, leader and people’s hero,” was established in Benxi city in Liaoning province (Stary, 1994:478). But the survival of the language, as G. Stary claimes, is limited now to a few isolated villages, and it is very difficult to predict its future (Stary, 1994:477)19. To conclude this very short survey of the history of the Manchus it is fair to say that the Manchu traditions, customs and folklore are more active now than the Manchu language itself. In the course of time the Manchus became one of the contemporary minorities in China and they reveal the signs of a resurgent identity at present.
4. The Manchu Language and its Dialects Long before the Manchu script came into being there existed a variety of dialects and subdialects which were more or less in correspondance with the multitude of Manchu clans and tribes. Literary Manchu took shape at the latest by the end of the sixteenth century and was based on the Southern dialect. According to the recent studies it was the language of the tribes which originally supported the first Manchu ruler, Nurhaci (Norman, 1974:159). In the course of time the Manchus who spoke the Southern dia lect assimilated with the Chinese and have lost their vernacular. Nowadays the Southern Manchu dialect can be reconstructed only on the basis of the old literary texts. Unfortunately, they have not yet been fully investigated. 19
In the People’s Republic of China three scientific journals entirely devoted to the Manchus, their history, language and culture, are now published on a regular basis. These are: Manyu yanjiu (“Manchu language research,” Harbin), Manzu yanjiu (“Manchu research,” Shenyang) and Manzu wenxue (“Manchu literature,” Dandong). Since 1992 “Manchu Research Institute” of the Academy of Social Science began publishing a new series dedicated to the Manchus. Taipei “Manchu Association” publishes the journal Manzu wenhua (“Manchu culture”) (Stary, 1994:477).
gorel-1.p65
30
4/17/02, 8:56 AM
31
prolegomena to manchu studies
Spoken Manchu is represented nowadays by the language of the Sibe, a Manchurian tribe. There are reasons to believe that Sibe was not a direct descendant of Literary Manchu, and at the time the latter was formed (on the threshold of the seventeenth century), the Sibes were not members of the confederation of Manchu tribes ruled by Nurhaci. However, according to a number of scholars, the language of the Sibe is very close to Manchu, and may properly be called a Manchu dialect (Norman, 1974:160). Besides, among Chinese scholars, who currently devote much attention to studying Sibe, there are differing opinions as to whether the primordial Sibe tongue was Manchu. Some of these scholars (e.g., An Jun, 1985), consider the Sibe language to be a successor of the Manchu language, whereas others (e.g., Li Shulan, 1983) hold the view that it is a separate language. It should be noted that nowadays the government of the People’s Republic of China has recognized Sibe as a distinct language which the Sibe people, one of the ethnic minorities of China, still speak. 4.1. The Manchus and the Sibes For a number of historical reasons the Manchu tribe Sibe (Chinese Xibo) resides nowadays in two regions of China, the North-east (Dongbei) and Xinjiang (Sinkiang). The majority of the Sibes inhabit Dongbei, the provinces of Liaoning, Heilongjiang and Jilin (53,560 people), but they have almost lost their ancestral language. In the Liaoning province whose capital city is Shenyang (Mukden), there are eight Manchu autono mous counties, but no Sibe speakers reported. In the Heilongjiang province there are a few villages where some Sibes, mainly old, still speak their native language. The Xinjiang Sibe, more than 27,000 strong, are concentrated mainly in the North, in the “Sibe-Cabcal Autonomous County,” instituted in 1954 along the south bank of the Ili river, in Xinjiang. More than 17 thousand Sibes (17,362 people out of 27,364) live there (Li Shulan & Zhong Qian, 1986:1-2). They have not only retained their ancestral language since the resettlement from Dongbei to Xinjiang in 1764-1765, but they have also been developing spoken Sibe and its written form, which is based on the Manchu syllabic script. Bearing in mind that the original homelands of the Manchu people
gorel-1.p65
31
4/17/02, 8:56 AM
32
part one
were the territories of Aigun and Fuyu, many questions naturally arise. How did the Sibes find their way to Xinjiang? What was the reason for their resettlement in that region? Furthermore, what is the origin of the language of the Sibe and what connection does it have with literary Manchu? However, very little is known about the origins of the Manchu tribe Sibe and their early history. It is known that the tribe Sibe was not a member of the Manchu confederation at the time when Nurhaci began his campaigns for tribal unification in the late sixteenth century (Norman, 1974:15960). S.M. Shirokogoroff mentioned a folk tradition according to which the Sibes were remnants of the Jin Jurchens (Shirokogoroff, 1924:1735). According to Crossley, there is no historical evidence for such an opinion. She maintains that the Sibe were conquered by the Man chus in the 1640s and then incorporated into Nurhaci’s federation. They are traditionally called “New Manchus” (Crossley, 1990:241; 1997:213). On the basis of her analysis of surviving Manchu manuscripts and Chinese records such as the dynastic chronicles of Liao (907-1125), published by the Manchus in 1647, in Chinese and Manchu, Tang (618-907), the earlier chronicles of Beishi (386-581) and Weishu (the History of the Yuanwei Dynasty, 386-535), the Russian scholar E.P. Lebedeva has suggested that the Sibe originated from the Southern Shiwei tribe which is a separate Tungus-Manchu branch of the Shiwei people (Lebedeva & Gorelova, 1994:14-5). In the History of the Liao Dynasty, the tribe is referred to as the Siwei. The problem of the origins of the Shiwei people is a matter of debate. According to A. Kessler, the Shiwei were a people of very ancient origins. Some historians think that they shifted to northeastern China from the Henan province at the end of the Xia period (2050 to 1650 BC), escaping the Shang conquest. Based on evidence from Chinese records, Kessler maintains that by the Tang dynastic era, there were as many as nine tribes which had the name Shiwei and lived in different territories in northeastern China. The Tang court established a Shiwei Area Commander to supervise these peoples (Kessler, 1993:145-6). Most Chinese historians belive that the Mongols were the descen dants of the Shiwei, and a tribe called Mengwu Shiwei, who origi nally lived on northeastern China, could be considered as the an cient ancestor of the early Mongols. Lin Gan maintains that a separate branch of the Shiwei tribes of the Eastern Hu line had given
gorel-1.p65
32
4/17/02, 8:56 AM
33
prolegomena to manchu studies
rise to the early Mongols20. Zhao Yue also thinks that the Shiwei, the term meaning “forest lands” in Mongolian, were ancestors of the early Mongols. He points out that in ancient times the Shiwei lived in the forests of the Hulunbeier area in northeastern Mongolia (Kessler, 1993:145-6). The Russian scholar N.Ja. Bichurin, based on the Beishi chronicle, maintained that the Shiwei people divided into five territorial groups (aimak), the Southern Shiwei, the Northern Shiwei, Bo Shiwei, Da Shiwei and Shengmo Shiwei. These groups differed one from an other in their occupations, the clothes they wore and the dwellings they lived in. All Shiwei’s aimak divided into generations (probably, clan groups). Thus, for instance, the Northern Shiwei, according to N.Ja. Bichurin, divided into nine generations (Bichurin, 1950:77). The Southern Shiwei called their leader “yumofo mandu”. The word “yumofo” contains the word “mofo” which corresponds to the Manchu word mafa “grandfather,” ”ancestor,” “old man” and to the Nanai word mapa “old man” (honorific vocative for an elder). The Southern Shiwei people were engaged in agriculture and cattle breeding. They lived in settlements similar to small towns. As means of conveyance they used carts and boats made from animal hides. When nomadizing, they built shelters using branches and covered them with grass. They traded with Gaoli, the Korean king dom, from whom they obtained ironware (Bichurin, 1950:77-8). The Russian scholar P.N. Menshikov, based on the histories of the Tang and the Yuanwei dynasties, also pointed out that there were several different tribes with the name Shiwei living in the Hulunbeier area and in the territories south of it. They spoke the language similar to that spoken by other peoples living in Manchuria, that is Tungusic. To the above-mentioned groups, he added the Lodan Shiwei and Menggu Shiwei who lived along the left and right banks of the Amur river, from the outfall of the Zeya river to that of the Songari river. From the mouth of the Songari, downstream of the Amur river, were located the lands of the Heishui Mohe (Menshikov, 1917:10). According to “Beishi”, as well as Yuanwei and Tang records, from the fourth to seventh centuries, the Shiwei people were governed by the Turks. 20 The Eastern Hu was the name for the Hu peoples who lived in southeastern Inner Mongolia during the Zhou period. According to A. Kessler, the peoples, living along China’s northeastern boundaries, were normally called the Di and the Hu in pre-Qin Chinese records (Kessler, 1993:37).
gorel-1.p65
33
4/17/02, 8:56 AM
34
part one
I. Dobrolovskii who studied the history of Heilongjiang province also believed that the Shiwei people resided in the Hulunbeier area. According to him, the lands of the Shiwei were located between the Heishui Mohe’s territories in the east and the Huns Tujue’s lands in the west, and in the south they bordered the Qitans (Dobrolovskii, 1908:34). Based on the Sui dynastic chronicle (581 to 617 AD), Dobrolovskii pointed out that the Shiwei people divided into five tribes, the Nan (Southern) Shiwei, the Bei (Northern) Shiwei, the Bo Shiwei, the Shengmo Shiwei, and the Da (Great) Shiwei (Dobro lovskii, 1908:34). It is probable that the general name “Shiwei” was applied to all these tribes despite their originally being of different ethnic back grounds. The majority of them were of Tungus-Manchu ethnic origin, but others among them were of Mongol origin, like for ex ample the ancient Qitans or the contemporary Dagurs. The Southern Shiwei lived in the territories that had previously been inhabited by the ancient Tungus state, the Fuyu, which was destroyed by the Xianbei people in 285 AD21. Fuyu was known for its developed farming. The Sibe who were located near Qiqihar, sowed wheat and millet, bred cattle, horses and pigs. Hunting and fishing were their traditional activities. Therefore it is considered that the Sibes not only inherited territory from this state (the region of Qiqihar and the area around the rivers Nomin and Nemer, tribu taries of the Nonni) but they took over from them agriculture as well (Lebedeva & Gorelova, 1994:16). According to Lebedeva, during certain historical periods the Sibe were overrun by other peoples. From the first century BC to the end of the third century AD, they were governed by the Koreans and the Chinese. From the end of the third, and probably throughout the fourth century, they were subjugated by the Xianbei people. The Northern Turks reduced them to a subordinate position from the 21 There are several schools of thought on the origins of the Fuyu people. According V.Ye. Larichev, they can be considered as people of Korean descend. During the Han period (206 B.C. to 220 A.D.), the Sushens had very complex relations with the Fuyu, their neighbours in south and southeast, to whom they were forced to pay a tribute. By 220 A.D., the Sushens were so hardly oppressed by the Fuyu that they led a rebellion against them. Before 226 A.D., the Sushens put an end to Fuyu’s dominion over them (Larichev, 1998:48-9).The Xianbei people were first mentioned in Chinese records as Eastern Hu affiliate tribes. In 386 the Tuoba Xianbei established the Northern Wei dynasty which lasted till 535 A.D.
gorel-1.p65
34
4/17/02, 8:56 AM
35
prolegomena to manchu studies
fifth century to the beginning of the tenth century. Subsequently, they were absorbed into the Qitan Empire of Liao (in the tenth century) and then of the Jurchen Empire (in the twelfth century), until by the thirteenth century they came under the dominion of the Qorcin Mongols (Lebedeva, 1985:24). In the mid seventeenth century a considerable part of the Sibe population actually resided in the regions near Bodune and Qiqihar. That period in the history of China and contiguous peoples was characterized, firstly, by the Manchus’ seizure of power and, sec ondly, by vigorous advance of the Russian Cossack detached forces to Amur where they were forced to fight a battle against the Manchus. Therefore the formerly neglected regions of the Heilongjiang prov ince attracted the Qing authority’s attention. Fortifications and roads began to be built, and local population was intensively enlisted into the Banner forces. Among that population the Sibe people were quite often mentioned. Brief but very important data on the Sibe’s history can be found in A. Rudakov’s book which is a translation of the Chinese manu script “Jilin tongzhi” (“History of the Jilin province”) compiled by Chinese scholars in the 1890s, with the translator’s additions. Ac cording to “Jilin tongzhi,” the region of Bodune was the main place of the Sibe’s ancestral home. By 1693 there were sixteen Sibe com panies in Jilin, and in 1693 another thirty Sibe companies were formed in Bodune which were then transformed to Mukden in 1700. After that, only those Sibes who were in the service of local Mon golian princes, remained in Bodune (Rudakov, 1903). The information concerning earlier history of the Sibe is very scant. According to “Jilin tongzhi,” there were two large aimak, the Sibe and Gåaleca, among those nine generations, Qorcin, Yehe etc., which declared war on the Manchus during the Taizu’s emperorship. In 1619 the tribes located towards South of the rivers Nonni and Ula, acknowledged themselves subordinate to the Manchus. Thus, the Sibe and Gåaleca (Gåwalca) were assigned to the Mongolian ban ner very early (Rudakov, 1903:11). It is very likely that the Sibe and Gåaleca, after they had been subjected by the Mongols, joined the union of the Mongol tribes and together with them fought against Nurhaci. Then, having been subjugated by the Manchus, they were assigned to the Mongolian banners.
gorel-1.p65
35
4/17/02, 8:56 AM
36
part one
We have no information about the time when the Sibes became vassals of the Qorcin Mongols. It probably happened when the Mongols started occupying the regions of the Songari and Nonni rivers and adjoining places. The Russian scholar P.N. Menshikov believed that the Qorcin Mongols belonged to the Mongols of the Jerim Seim who moved to these regions from the banks of the Argun and Haidar rivers and Dalai Nor lake in 1438 after they had been defeated by the Oirats. The regions of Qiqihar and Bodune were the original places of the Sibe, whereas the Mongols were newcom ers (Menshikov, 1917). There is more historical evidence that over a long period the Sibe inhabited the Bodune and Qiqihar regions, their ancient homeland. The Manchu manuscript entitled “Sahaliyan ula vajimbuhe baitai dangse” (the Journal of Completed Documents Concerning the Amur Re gion) contains documents dating from the turn of the eighteenth century (the twelfth of the first month of the thirty eighth year of Kangxi’s reign, i.e., in 1700) which infer that the Sibe were ransomed from Mongolian slavery at the end of the seventeenth century by the Emperor Kangxi. At that time the Sibe were settled in the re gion of Bodune along the rivers Liaohe, Shira Muren and Yangshailu and in Ula territory. The same documents show that by the end of the seventeenth century two large groups of Sibes existed. The first group inhabited the region of Qiqihar, the other lived near Bodune (SU). Later, in 1700-1701, the Qiqihar Sibe (numbering nearly 20,000) were resettled in the Huhehaote (Huhu hoton) region, and their fur ther destiny is still unknown. The Bodune Sibes (nearly 36,000) were transferred to the Mukden region. N.N. Krotkov mentions in his memoirs that in the thirty fifth year of the Emperor Kangxi’s reign (1697) the Manchu tribe Hoifan (Hoifa) rebelled against the Qing authorities and was exterminated by the regular forces, and in the forty-first year of the Emperor Kangxi’s rule (1703) the Manchu tribe Ula ceased to exist (Krotkov, 1911-1912:117-37). The rebellions of the tribes Hoifan and Ula took place at the time when the Sibe’s resettlement had been proceed ing, and the coincidence between these events could hardly be con sidered as accidental. In 1764 (the thirtieth year of Qianlong’s reign) nearly half of the Mukden Sibes, about 18,000 of them, were transferred to the Ili
gorel-1.p65
36
4/17/02, 8:56 AM
37
prolegomena to manchu studies
Valley of Xinjiang to colonize Jungaria, where their descendants still live today. It is as yet unclear why the Sibes were chosen for this purpose unless it was their opposition to the ruling Manchu Dynasty (Lebedeva & Gorelova, 1994:10-1). In the document number seven of the above-mentioned collection there is information about disor der among the Sibes. It is also mentioned that the Qiqihar Sibe companies’ commanders and their officers were removed from their posts. During their transfer from Mukden to the Ili Valley, the Sibes were convoyed by Manchu regular forces of eight hundred officers and men. All these facts corroborate the compulsory nature of the Sibe’s transfer to Xinjiang (SU). Conquered in 1755, Xinjiang was a place where bannermen were installed as agents of occupation, in order to keep peoples of varie gated culture obedient, and even in the nineteenth century it was rather difficult to keep these peoples of different ethnicities and different languages pacified (Crossley, 1989:13, 18-9). Even at the end of the ninteenth century Xinjiang was a place where bannermen were sent for their faults, a place for punishment. In Lao She’s novel we can find indirect confirmation of this fact. One of the person ages feared that if he didn’t act according to the edicts of the au thorities, “they would take his name off the Bannerman register and send him to Xinjiang or Yunnan” (Lao She, 1982:137). Later on, in accordance with the Eight Banners’ Organization of the Manchu army, the Sibes were incorporated into eight compa nies and assigned to seven military colonies (the first and the third companies shared the same military colony) together with the Solons and Dagurs, who had previously been absorbed into the Manchu banners. Due to this fact, the Sibes increased in number. By con trast, in the second part of the nineteenth century they declined in number as a result of being forced to participate in suppressing the Dungan-Taranchi rebellion in 1864-1871. During this armed rising the Sibes’ villages were destroyed and they sustained heavy losses. In Xinjiang the Sibes were engaged only in agriculture. They sowed wheat, barley and millet, cultivated onions, garlic, aubergine, cucumbers, and other vegetables and grew corn, tobacco-plant and poppy as well. As food they used farm produce, fish, and very rarely meat. Their clothes did not differ from those of the Chinese. Their standard of living in old China was very low. The Sibe people professed shamanism and Buddhism. There was
gorel-1.p65
37
4/17/02, 8:56 AM
38
part one
a Buddhist monastery in the military colony where the fifth com pany was located. Besides, every colony had its own Buddhist shrine (Krotkov, manuscript). Having settled in Jungaria, the Sibe people found themselves in a peculiar situation. On the one hand, their remoteness weakened the influence of Chinese culture and language. On the other hand, living among linguistically and ethnically different peoples fostered a consciousness of their national character. The language of the Sibe in its turn was subjected to direct influence from the languages of the other Xinjiang peoples, primarily the Kazakhs, Uigurs, Mongols, as well as the Chinese and Russians, but at the same time it revealed a tendency towards self-preservation and development. 4.2. Contemporary Life of the Sibes in Xinjiang, One of the Most Multinational Regions of China After their resettlement to Xinjiang in 1766, the Sibes were forced to learn to live as one ethnicity among peoples of different cultural and ethnic affiliation. They kept their distance even from the Manchus proper who had taken the dominating position in the region. During a long period of time until the first years of the Republic the Sibes retained their language, customs and religion. After the foundation of the People’s Republic of China the Sibes continued living in Xinjiang as one of the minorities, preserving their traditions and developing their language. This situation seems es pecially remarkable if we bear in mind that the authorities’ report on the situation of minorities in Manchuria claimed in 1952 that there was nobody who spoke or read in Manchu in the Northeast (Stary, 1994:476). Before the Cultural Revolution in China in the “Sibe-Cabcal Autonomous County” teaching in primary school was carried out in the Sibe language. At present in the Xinjiang province there are eight schools where pupils learn Sibe. In spite of the fact that teaching in primary school is conducted in Chinese, lessons on the Sibe lan guage are obligatory. How long the pupils can attend the Sibe lan guage classes depends on the number of teachers who can teach Sibe. For the first time the Sibe language has begun to be studied in a higher educational institute. In Ili there is a Pedagogical Institute where a group of students studies Sibe. In Cabucar there exists a Pedagogical Institute of the Sibe language. This educational insti
gorel-1.p65
38
4/17/02, 8:56 AM
39
prolegomena to manchu studies
tute trains students for teaching the Sibe language in the national school. The People’s Publishing House was organized in Urumqi in 1953. It has concentrated its editorial activity on printing party documents, belles-lettres and schoolbooks in Sibe. There was an interruption during the Cultural Revolution, but after that the House resumed publishing with great enthusiasm in the 1980’s. G. Stary has given a survey of the editorial activity of the People’s Publishing House of Urumqi from 1953 to our time in his Foreword to “A Catalogue of Sibe-Manchu publications” written by the Sibe scholar Jin Ning. This catalogue contains 285 entries and includes the works in Sibe, pub lished during the period from 1954 to 1989 (Jin Ning, 1989:V-VI). The newspaper “Ice Banjin” (“The New Life”) has been printed in Ili since 1946. In 1972 the title of this newspaper was changed to “Cabcal Serkin” (“The Cabucar Newspaper”). Under the name of “Cabcal serkin” this newspaper is now regularly published in Xinjiang. In 1980 in Cabucar “The Society of the Sibe language” was organized to improve and develop the native language of the Sibe people. G. Stary who visited Cabcal Autonomous Country, Xinjiang, in 1991, described some elements of Sibe-Manchu culture that have been preserved carefully up to now. Every year, on the eighteenth day of the fourth month of the lunar calendar, the Sibes gather to commemorate their coming to Xinjiang from Manchuria to protect newly conquered territories. They celebrate a great date of the westward migration of the Sibe people to the Cabcal self-administered county with a folk perfomance of a “shaman dance.” In Cabcal town all signs in the main street were written in Sibe, followed by Chinese. G. Stary took some pictures of the main street of Cabcal town, a typical Sibe house in the “Village of the Fifth Company” and the Sibe-language Middle school in the “Village of the Seventh Company”. He also photographed some historical places: a tombstone stele in the cemetery of the Fifth Company, the Lama temple of the Fifth Company, the remains of Dolantu Karun, one of the sentry-posts where the Sibes stood guard, shaman ritual objects which are kept now at Aisin Sheri, etc. (Stary, 1995:95-117). A. Pozzi who visited “Sibe-Cabcal Autonomous County” in 1991, collected extremely interesting evidence of the religious traditions of the Sibe people whose ancestral religion was shamanism. Her article is devoted to the description of the cults of two divinities that
gorel-1.p65
39
4/17/02, 8:56 AM
40
part one
belong to the pantheon of the Sibes. One of them, the female di vinity Siri (or Sirin) mama, protects the continuity of a family and goes back to the matriarchal society. Siri mama is represented by a silk cord with many different wooden or paper objects attached to it. The other, the male divinity Hairkan (or Harikan) mafa, protects domestic animals and goes back to ancient times when the Sibes began to live in the patriarchal society. Hairkan mafa is represented by a wooden box with paper talismans, horse’s heads made from fabric, etc., inside it, and the box very often contains an image of the divinity, a middle-aged bearded man in a sitting position. For the older generation these two divinities have a religious meaning while for young people they became a part of everyday life (Pozzi, 1993). As it appears from the above, the Sibes, being genetically and linguistically close to the Manchus, represent a separate branch of the Tungusic world. Although their history was closely related with the history of the Manchus, the Sibes have their own distinctive fate which can be traced to their origin. They have not only managed to preserve their specific ethnic features, but in recent decades they have been successful at promoting the development of their own original and distinctive culture. Analyzing the histories of the Manchus and the Sibes, one can make a good case for claiming a very rare instance in the history when a separate branch of a language family, being the sole survi vor, reveals a tendency to supplant the latter on the world arena. In certain sense the history of the Sibes can be considered as the continuation of the history of the Manchus. 4.3. The Sibe Dialect in Relation to Literary Manchu For more than 200 years, since the resettlement of the Sibes in Xinjiang, both the spoken and written forms of the Sibe language have been changing and developing. The Sibe dialect was recorded as a separate language for the first time in the 1860’s, when the Russian scholar V.V. Radlov wrote down the Sibe text “Sidi Kur” from a native speaker, Bite. It seems probable that this text reflects the language of educated members of Sibe society of that time. Educated Sibes maintained the purity of Literary Manchu to a much greater degree than the Manchus who lived in Beijing (Peking) and Jilin and Heilongjiang provinces. The
gorel-1.p65
40
4/17/02, 8:56 AM
41
prolegomena to manchu studies
language of this text differs slightly from Literary Manchu, the main distinctions lying in the sphere of phonetics (Lebedeva & Gorelova, 1994). The second recording of the Sibe dialect took place at the begin ning of the twentieth century, when F.V. Muromskii, another Rus sian scholar, visited the Ili Valley in Xinjiang twice, in 1906 and 1907-1908, and made a large number of recordings in Sibe. He paid special attention to collecting samples of living, colloquial, language; these texts were later published by the Polish scholar St. KaÑuîiÔski. This version of Sibe differs from Literary Manchu to a greater ex tent than Radlov’s recordings, but the distinctions are again of a phonetic nature; the main morphological forms simply having dif ferent phonetic shape (KaÑuîiÔskii, 1977). These two text corpora reflect two distinct forms of the Sibe dia lect, the written and the spoken ones, respectively. Among other publications directly concerning the language of the Sibe, “A Classified Dictionary of Spoken Manchu” compiled by K. Yamamoto, should be mentioned. The dictionary published after K. Yamamoto’s death by S. Hattori, is prefaced by their joint article discussing vocalic and consonantal inventories of the Sibe dialect (Yamamoto, 1969). As mentioned in the preface to the dictionary, the lexical material was recorded from an informant who was a native Sibe and who spoke two dialects of the Manchu language without differentiating between them. Then, a sketch of Sibe morphology with a number of tables representing vocalic and consonantal inven tories and basic morphological markers, was published by J. Norman. This linguistic material was collected by him in Taiwan in 1960-1970 (Norman, 1974). It would probably be wrong trying to reconstruct a kind of uni fied Sibe dialect on the basis of such different language samples. On the other hand analyzing the differences between these versions is also of certain interest. An attempt to define the differences in phonet ics and morphology has been presented in previous works by the author (Gorelova, 1986; Lebedeva & Gorelova, 1994:21-103). In China itself a great number of works devoted to the language of the Sibe were published. The most notable of them belong to Chinese scholars An Jun, Li Shulan, Zhong Qian, Wang Qingfeng, etc. Li Shulan who is the best-known researcher of Spoken Sibe, published a series of articles, the language base of which were samples of native Sibe speech collected mostly in Xinjiang (Li Shulan, 1983).
gorel-1.p65
41
4/17/02, 8:56 AM
42
part one
She also wrote several considerable works together with Zhong Qian and Wang Qingfeng. These publications present the description and a valuable vocabulary of Spoken Sibe (Li Shulan & Zhong Qian, 1986; Li Shulan & Zhong Qian & Wang Qingfeng, 1984). A recent contribution concerning colloquial Sibe has come from the Sibe scholar Jin Ning whose publications contain a collection of every day vocabulary (Jin Ning, 1993). During the past decade Chinese scholars have described the dif ferences between written and spoken forms of Sibe from Literary Manchu in the belief that written Manchu reflects to a certain de gree spoken Manchu of the 16th-17th centuries. The following is a list of the main distinctions of the written form of Sibe from Literary Manchu, as presented in Li Shulan’s paper (Li Shulan, 1983:298-306): Sibe script, which is based on Manchu script, underwent several changes. The style of writing some letters in certain positions was modified. Some seldom-used syllables were eliminated, so that 13 out of the 131 syllables occurring at word initial position in written Manchu are no longer used in contemporary Sibe. By contrast, some new syllables (wi, wo, wu) were added. The most significant changes, primarily in vocabulary, resulted from contact with neighbouring languages, with loan words com ing from Chinese, Uigur, Kazakh, Mongolian and Russian. Having resettled in Xinjiang in 1764, the Sibe people borrowed a number of words from local Chinese and other peoples. For example, words were borrowed from Uigur: pochi “boaster;” parang “words;” ketman “tool for digging;” bazar “market;” kavab “barbecued meat;” namas “an Islamic feast”. A number of words came from Kazakh: kemes “Mongolian tea;” kestao “pasture;” baige “horserace,” “races”. Some words came into Sibe from Russian: baston “a kind of textile,” masina “sewing-machine;” kongsol “consul;” miter “kilometer;” pamidor “to mato”. After the Xinhai Revolution (xinhai geming) some new terms ap peared in the Sibe language to designate new realities and notions. Most of them were borrowed from the Chinese language. This process of borrowing from Chinese is still underway. Representa tive examples include: geming “revolution;” fangzhen “guiding prin ciple,” “policy;” jiti “collective (body);” jingji “economy;” zhuxi “chair man,” “chairperson;” zongli “prime minister;” shuji “secretary;” daikuan “loan,” “credit;” chuna “cashier;” gongfen “centimeter,” “gram”.
gorel-1.p65
42
4/17/02, 8:56 AM
43
prolegomena to manchu studies
The grammatical changes were not great, being confined mostly to a slightly different usage of a few markers among those common to Sibe and Manchu and to formation of new markers on the basis of the old ones. For instance, based on a construction conveying simultaneity and consisting of the imperfect converb (marker -me) and the perfect finite form of the verb ili- “to stand,” the new marker -mahabi was formed with the same meaning. The spoken form of Sibe differs from Literary (Written) Manchu to a much greater extent than the written form. The most striking discrepancies can be observed in the area of phonetics, where the reduction of vowels as well as consonants is characteristic of Spo ken Sibe. Some instances in which this reduction appears are: 1. Manchu unstressed (syllabic) vowel a is reduced to [�]: ma. amtan [amtan] – sib. [amt�n] “delicious,” “tasty,” ma. galman [�alman] – sib. [�alm�n] “mosquito,” ma. tatambi [tatambi] – sib. [tat�m] “to draw,” “to pull.” 2. The diphthongs occurring in the initial position become simple vowels: ma. aiha [ai�a] – sib. [��] “pearl,” ma. eifu [�ifu] – sib. [if] “grave.” 3. Manchu fortis consonants b and f are realised as the lenis consonant v: ma. gebu [g�bu] – sib. [g�v] “name,” ma. abalambi [abalambi] – sib. [av�l�m] “to hunt.” Another characteristic of Sibe phonetics is the dropping of vowels. 1. Unstressed vowels at the end of Manchu words are usually omitted in their Sibe counterparts: ma. baita [baita] – sib. [bait] “affair,” “matter,” ma. abdaha [abda�a] – sib. [af�] “leaf.” As a result of the loss of vowels, a new type of syllable has appeared in Spoken Sibe – that with V + C + C + C: ma. antaha [anta�a]— sib. [ant�] “a guest.” 2. Not only vowels are often lost, whole syllables may disappear: ma. mudalin [mudalin] – sib. [m�dan] “a bend of a river.” Some syllables are shortened to the extent that the morphological
gorel-1.p65
43
4/17/02, 8:56 AM
44
part one
markers they designate radically change their phonetic shape. Thus, the Manchu syllable -ngga is shortened to [�] in Spoken Sibe: ma. aldangga [aldangga] – sib. [ald��] “distant,” “far.” The Manchu marker of the imperfect finite form -mbi transforms in Spoken Sibe into [m]: ma. deyembi [d�j�mbi] – sib. [d�j�m] “to fly,” ma. banjimbi [band�imbi] – sib. [band�im] “to grow.” Some new vowels which are absent in Literary Manchu have appeared in spoken Sibe. These vowels are [�] and [�]. They took their origin not only from the diphthongs but also from the vowels a and o occurring at the initial position and followed by the vowel i: ma. adaki [adaki] – sib. [�dki] “neighbor,” ma. alin [alin] – sib. [�lin] “mountain,” ma. dobi [dobi] – sib. [d�vi] “fox,” ma. omimbi [omimbi] – sib. [�mim] “to drink.” Finally any brief description of Sibe phonetics, must note the phenomenon of full assimilation: ma. edun [�dun] – sib. [udun] “wind,” ma. muten [mut�n] – sib. [mutun] “skill,” ma. tuhembi [tu��mbi] – sib. [tu�um] “to fall.” As a result of such huge changes in phonetics, the previous state of vowel harmony in Spoken Sibe has been drastically destroyed. Quite possibly that one can speak about a new kind of sinharmonism in Spoken Sibe (see examples cited above). A characteristic feature of Radlov’s texts is vowel harmony, which is essentially the same as in Literary Manchu. J. Norman indicates the alternation of the per fect participle suffixes -x� and -h� as the last vestige of vowel har mony in the dialect described: -x� only occurs with front vocalic stems having the vowels i, u and �; -h� occurs with verbal stems that contain one of the back vowels u, o, a or � (an umlauted variant of a) (Norman, 1974:170). With the vowel harmony having totally disappeared, the consonants retain some harmonic features (x, h). The vocabulary and morphology of Spoken Sibe have also un dergone modifications as compared to its written form, but these are not very significant. The most important grammatical feature of Spoken Sibe in contrast to Literary Manchu is the category of per sonal possession. The presence of this category brings together Sibe and the Tungusic languages and hence is of typological significance.
gorel-1.p65
44
4/17/02, 8:56 AM
45
prolegomena to manchu studies
The most commonly used marker is that of the third person [ni�]: ma. tere i eniye jihebi “his mother came”—sib. [t�r�j �ni�-ni� d�i��i] “his mother-his came”. There are also markers of the first and sec ond persons, but they are rarely used: sib. [mini ma�al-mini] “my hat-my;” [�ini uv �uturi-�ini] “your fortune-your”. Li Shulan suggested that the category of personal possession ex isted in Spoken Sibe before the resettlement of the Sibes to Xinjiang and did not appear as a result of the subsequent changes (Li Shulan, 1983:305). To conclude these brief observations of the distinctions of Spo ken Sibe from Literary Manchu it must be said that Sibe, as the lan guage of a marginal nation, is not in a stationary state. It is devel oping and altering. The changes in phonetics are so great that Sibe requires a status of a distinct language of Tungus-Manchu origin.
gorel-1.p65
45
4/17/02, 8:56 AM
This page intentionally left blank
47
manchu script
PART TWO
MANCHU SCRIPT
gorel-1.p65
47
4/17/02, 8:56 AM
This page intentionally left blank
49
manchu script
PART TWO
MANCHU SCRIPT 1. The History of Writing the Manchu Language When the Manchu people emerged as a political force in the Oriental world in the late sixteenth century, they used at first the Mongol language and Mongolian script in official court documents. As it is well known, the syllabic Mongolian script was based upon the Uigur one which in its turn, derived from Sogdian and, origi nally, via Syriac, from Aramaic. Derived from the ideographic script of ancient Egypt around 1000 BC, Aramaic was phonetic in nature which meant that it used a finite set of graphic elements to repre sent the sound system of a language. The Aramaic script was pre dominant among peoples of the eastern Mediterranean, including Syria. Bearing in mind that the Syriac script was horizontal which is meant that the language was written from left to right there arises an important question. Why did Turkic Uigurs when adapting the Semitic Syriac script, change its direction, writing their own language vertically, from top to bottom? According to earlier and recent specialists, this was done by rotating the script by ninety degrees counterclockwise in order to imitate the Chinese script which was very prestigious in Central Asia at that time (Zakharov, 1879:47-8; Crossley, 1997:35-6). It is important to mention that medieval Jurchens used a script of another type in their communications with the Chinese and Korean courts in Ming times. That script was modelled on Qitan which combined some features of the Chinese ideographic script with some phonetic elements borrowed from Uigur. As it turned out, due to its phonetic elements the arrangement of the Qitan script was better adapted for rendering both the Qitan and Jurchen languages whose grammatical structures were Altaic by nature, although Qitan was much closer to Mongolian. Jurchens used that script until the fall of their Jin empire in 1234, and even much longer in their of ficial communications. Then in the late sixteenth century the Jianzhou Jurchens from whom the Later Jin khanate arose, recog
gorel-1.p65
49
4/17/02, 8:56 AM
50
part two
nized that nobody in garrisons could read old Jurchen script, and even headmen wanted to communicate with the Chinese and Ko reans in Mongolian (Crossley, 1997:37-8). Because the Mongolian script did not transcribe all sounds of the Manchu language, the Manchus needed a more versatile alphabet. Therefore, during the reign of Nurhaci, the process of adaptation of the Mongolian script to the Manchu language commenced. In 1599 Nurhaci commissioned two Manchu learned men, Erdeni-bakài and the minister Gagai-jarguci, to devise a new script1. They adapted the Mongolian script to the Manchu sounds without noticeable changes, and the Manchus used this alphabet until 1632. The ear liest period of the developing of the Manchu writing dated from reign periods Tianming (1618-1626) and Tiancong (1627-1635). Although not many written monuments of that period survive, they are unique sources for the reasearch of the early history of the Manchus and their relations with neighbouring peoples (Pashkov, 1963:11-2). Later on, in 1632, the Manchu scholar Dahai essentially improved the Mongolian alphabet. He brought into use some diacritic signs, viz. dots and circles (ma. tongki and fuka correspondingly), and through that, in contrast with Mongolian polyphonic letters, every Manchu sound has a separate letter. The script itself came to be called “circled and dotted script” (Zakharov, 1879:1-2; Pashkov, 1963:12; Crossley, 1997:38). In spite of the fact that the Manchus themselves probably considered their script as syllabic, in fact it was writing in letters. Another problem had arisen. The script did not transcribe sounds of Chinese and Sanskrit words borrowed by the Manchus. There fore Dahai added ten graphemes, “tulergi hergen” in Manchu (liter ally “foreign (outer) letters”), to the Mongolian alphabet to render Chinese (mostly in Chinese names, titles and offices) and Sanskrit sounds which were absent in Manchu. Investigating also syllabic patterns of Manchu words, Dahai dis tributed all syllables, 1441 in number, among twelve classes. The Manchu writing system itself, described by Dahai as the twelve di visions of the syllabary, had got the name “Juwan juwe uju bithe” in 1 The word bakài, borrowed from Mongolian, means “a scholar,” “a learned man” (NL:24). See also Crossley who translates the word bakài as “a literate man” (ma. bakài < mo. baghshi < chin. boshi “an erudite”) (Crossley, 1997:38). The word jarguci is possibly borrowed from classic Mongolian (ma. jarguci < class. mo. j�ar?uc�i “an official”); see also the modern Mongolian word zarguc�i “a judge” and the Kalmyk word zargac� “a judge”).
gorel-1.p65
50
4/17/02, 8:56 AM
51
manchu script
Manchu (literally “twelve-heads-writing;” in Chinese “shi er zi tou”). It should be noticed, that practically not all syllables, indicated in tables representing the twelve divisions, existed in Manchu. A word must be said about the so-called Manchu “square-character script” (ma. fukjingga hergen) which was created in 1748. At first that style of script was devised to print the “Ode to Mukden” which was written by the Qianlong emperor. Then the square script was used for inscriptions on the seals which belonged to emperors, their wives and sons, to the ruling elite and commanders of the Imperial Army. Creating their square script, the Manchus tried to imitate both the ancient Chinese and the Mongolian scripts, devised exclusively for writing on seals. Presumably, the inscriptions on the Taizong emperor’s seal are examples of that kind of the Mongolian writing. The Manchus devised their square script in 32 variants, indicating which variant should be used and in what circumstances. Created for the Manchu nobility, that script never succeeded in replacing the Dahai’s alphabet (Pashkov, 1963:12, 14).
2. Graphic Structure of the Manchu Letters Although there is some evidence that the Manchus themselves were likely to perceive their own writing system as syllabic, and learned it as syllabary, in fact it became a writing in letters after Dahai carried out his reforms. Furthermore, all the letters were constructed using a small set of the following five basic elements: 1) the head, which is the beginning of a letter; 2) the circle; 3) the tooth (ma. a); 4) the connecting vertical line linking all syllables into a single word; 5) various additional elements which are represented by the so-called “tails” (ma. uncehen) and vertical, horizontal, and oblique line seg ments. These line segments, twelve in total, are drawn to the right or to the left of the connecting line. They may end with or without a hook, turned upwards or downwards. The head � is used as the beginning of a letter to constitute the following graphemes: a= , e , iE , o� , uK , å�E , n� , l � , m� , f . 2.
2 In real Manchu writing all these symbols would be rotated by ninety degrees clockwise.
gorel-1.p65
51
4/17/02, 8:56 AM
52
part two
The circle � is used to constitute the second half of the letters o �� , u �� and å �E � Slightly modified, it is used to constitute the first part of the velar and uvular counterparts of the letters k, g, h ( [k] � , [g] �� , [x] �� ; [q] � , [�] �� , [�] � ) and the letters b> , pF , t6 , , , d �� , �� . The tooth � is used to form the second part of the letter a �� at the beginning of a word, the letters a � and e � (when e is written like the letter a, but with a dot), when they occur in the middle of a word in combination with the letters n (na �� , ne �� ), k (ke �� ), g (ge � ), h (he ��� ), s (sa �� ), à (àa �= ), t (ta �� , te �� ), d (da6�= , de �� � ), l (la �� , le �� ), m (ma �= , me �A ), c (ca += , ce +A ), j ( ja �� , je �A ), r (in the middle of a word: ra �� , re �� ),� f (fa B� , fe �� ), v (w) (wa �� , we �� ), b (ba �� , be � ), p (pa � , pe � � ). The tooth is also used to constitute the uvular counterparts [q], [�], [�] of the letters k, g, h when they occur before the vowels a, o, and å in the middle of a word ( � , �� , � ). All additional line segments are of the following two types. Some of them, are obligatory elements of the graphemes which they are part of. There are nine of them: 1) the line segment with a hook turned upwards and written on the left is an essential element of the letter l � ; 2) the same line seg ment but with a hook turned downwards constitutes the letter m � ; 3) the oblique line segment on the left indicates the letter j in the beginning of a word ( � ) or the letter i in the middle of a word (� �� ), or that followed by the letter i (ii �� ) as well; it also repre sents the second part of the letter i at the beginning of a word ( �� ); 4) ending with a hook turned upwards, this oblique line in dicates the letter y � ; 5) crossed with a stroke, this oblique line indicates the letter r � ; 6) the oblique line segment with a hook turned downwards indicates the letter v (w) � ; 7) the same hook preceded with a stroke, which is drawn on the right and upwards, constitutes the letter f � ; 8) two line segments joined at an acute angle form the letter s and à (� , � ); 9) two line segments joined at right angles indicate the letter c � . Various final tails are used when a word ends in a certain letter. These tails are as follows. 1) The tail turned to the right is an element of the letter a when it occurs separately or after the letters n, k, g, h, s, à, t, d, l, m, c, j,
gorel-1.p65
52
4/17/02, 8:56 AM
53
manchu script
f, v (w) at the end of a word ( � ), in that very case the tail replaces the tooth which indicates the latter a in the middle of a word. 2) The same element is used to indicate the letter e when it is written like the letter a but with a dot on the left and stands at the end of a word ( �� ). 3) The tails turned to the left may be of two kinds according to their shape. The long tail � indicates the letter e when it stands alone ( �� ) or preceded by the letters k, g, h for velars [k] ��� , [g] ��, [x] ���� . The long tail also indicates the letter a after the let ters for the aspirated consonants k� �� , g� �� , h� �� as well as the letters a and e preceded by the letters b, p ( ��� , �� ; �� , �� � � )3. The long tail indicates the letter ng ( ��� ) as well. 4) Tails are short when they occur at the end of a word as ele ments of the letters i, o, u, and å ( � ; � ; � final, ��� isolated; �� final, ��� isolated). The short tail also indicates the letter r when the latter occurs at the end of a word ( �� ). The connecting line segment links all syllables into a single word and it is cut only by the velar counterparts of the letters k, g, h that is [k] � , [g] �� , [x] �� , or by the letters b � , p � , s � , à � , ng �� , for instance: abka ����� “sky,” “heaven;”� enggemu �������� “saddle;” ekàembi ������� —the imperfect form of the verb ekàe- “to hurry,” “to hasten;” babe ���� ; kara ����� “black (of animals);” unggihe K=CC�EC� =� —the perfect participle from the verb unggi- “to send,” “to dispatch;” obonggi �������� “bubble,” “foam.” In addition to these principal graphic elements, to distinguish certain sounds, Dahai introduced two diactiric signs, viz. a dot (ma. tongki) and a circle (ma. fuka). Thus, the dot placed on the right of a letter, is used to distinguish the letter e ( � , �� ) from the letter a ( � , � ) in the middle and at the end of a word, as well as the letter u ( �� , � , � ) from the letter o ( �� , � , � ) in the following syl lables: ne �� , nu �� , se �� , su �� , àe �� , àu �� , le �� , lu �� , me �� , mu �� , ce �� , cu �� , je �� , ju �� , re �� , ru �� , fe .A , fu �� , ve (we) �� , be � , bu �� , pe �� , pu � . A dot on the right is used also to distinguish the velar and uvular counterparts of the letter g that is [g] �� and [�] �� from those of the letters k and h 3
gorel-1.p65
The sign
h
is used to mark aspiration.
53
4/17/02, 8:56 AM
54
part two
([k] � , [q] � ; [x] �� , [�] � ) and the letter d ( �� ,� �� , �� , �� ) from the letter t ( � ,� � ,� � ,� � ). The dot placed on the left is used to mark the letter n (� ,� � ). The doubled dot on the left indicates the letter k after the letter rendering a vowel and before the letter rendering a consonant, for instance: akjan ������ “thunder,” sakda ����� � “old man.” The circle placed on the right is used to distinguish the velar and uvular counterparts of the letter h that is [x] �� and [�] � from those of the letters k and g ( [k] � , [q] � ; [g] �� , [�] �� ). The Manchu alphabet which is used today in a slightly modified form, includes 34 graphemes. Six of them are for vowels: a, e, i, o, u, å; 22 graphemes are used to render consonants: n, ng, [k], [g], [x], [q], [�], [�], b, p, s, à, t, d, l, m, c, j, r, y, f, v (w). Ten special letters were created to render the Chinese sounds which were absent in Manchu but borrowed by the Manchus with loan words: k�, g�, h�, ts\(c), ts (ci), dz (z), sy (si), z� (r), c\y (chi), jy (zhi). For transcribing certain Chinese sounds and syllables I use the symbols introduced by P.G. von Möllendorff in his Manchu Grammar (Möllendorff, 1892). In brackets the same sounds and syllables are also given in accordance with contemporary tradition (so-called pinyin system) accepted in the People’s Republic of China. Most graphemes have three variants of writing. Those are con ditioned by the position of a letter (initial, middle and final) within a word. Isolated letters, occurring out of both a word and a syllable, have their own specific modes of writing.
3. The Manchu Graphemes and the Way of their Writing Below are listed all Manchu letters, in most cases explaining the way of their writing in order to facilitate the readers’ understanding. As mentioned above, with a few exceptions which will be described further, most Manchu letters occur in three variants depending on their position in a word—at the beginning, in the middle or at the end. The obvious thing to begin with is the observation of the way of writing letters which render vowels. Subsequently graphemes which render consonants should be observed. Each letter should be repre sented by three variants—if they are (not all the letters have all of the three variants)—in the following order: at the beginning, in the
gorel-1.p65
54
4/17/02, 8:56 AM
55
manchu script
middle, and at the end of a word. The table of graphemes which render vowels also includes the fourth variant of the mode of their writing that is used when they occur alone, out of the graphic con text. Consonants never occur separately, out of either a syllable or a word. 3.1. The Manchu Graphemes Rendering Vowels (and the semivowel y) Each numbered line begins with the transliteration of a certain Manchu letter. This transliteration mostly coincides with the pho nological interpretation of the letter, but not in all cases. Each Manchu letter rendering a vowel is given here in four variants: When alone (isolated)
1) a
At the beginning of a word
In the middle of a word
At the end of a word
��
��
�
���
When alone, the letter a consists of the head, the tooth and the tail turned to the right. In the initial position the tail is replaced by another tooth. In the middle of a word the letter a is rendered by the tooth which in final position after consonants changes into the tail turned to the right, for example: alaha ������ —the form of the perfect participle of the verb ala- “to tell,” “to report.” Sometimes both the tooth in the middle of a word and the final tail can be mistaken for a shape of the letter n. To avoid a mistake, it should be taken into account that if two teeth are evenly apart, then it is a sign for the letter a; if one tooth is more distant from the others then it is a sign for the letter n, for example: acan ����� “har mony,” “concord,” “union,” “meeting;” encu ����� “different,” “other,” “strange.” 2) e
��
��
��
�� �� (after the letters for the velars [k], [g], [h]; for the consonants k�, g�, h�, b, p)
gorel-1.p65
55
4/17/02, 8:56 AM
56
part two
The isolated letter e consists of the head, the tooth and the long tail turned to the left. In the initial position the tail is replaced by a slightly longer connecting line segment. In the middle the letter e has a shape of the tooth with a dot on the right. In the final posi tion it consists only of a tail with a dot on the right, for example: eletele� �������� “until (it is) enough,” “in sufficient quantity.” Occurring after the letters p and b, the letter e is indicated by a dot placed on the right of these consonants, not of the letter e itself: benehe ������� —the form of the perfect participle of the verb bene“to send (away from the speaker),” “to deliver,” "to give as a gift.” 3) i
���
��
��
��
When alone, the letter i consists of the head, the tooth and the rounded short tail turned to the right. At the beginning of a word this tail is replaced by an oblique line segment drawn on the left of the letter. In the middle only this oblique line segment remains. At the end of a word the vowel i is indicated by the short tail drawn from the left to the right. Here are some examples: ili ���� —the form of the imperative of the verb ili- “to stand,” “to stop;” iliha ������ the form of the perfect participle of the verb ili- . Being the element of combinations which render the descending diphthongs ai, ei, oi, ui, or occurring after the letter i, this letter is written like two oblique line segments drawn on the left of the con necting line. In other words, the letter i is doubled when it is used to render the element i � in the middle of a word: (�� ). The first of these oblique line segments may be replaced by the tooth ( �� ), for ex ample: eimembi �������� —the imperfect finite form of the verb eime“to abhor,” “to detest,” “to find unpleasant.” At the end of a word it has a shape of a short tail turned to the right (�� ), for example: abkai ������ < abka “sky,” “heaven” + i (genitive). The letter for the ele ment i � occurs separately only as the form of the genitive case: �� . ��
4) o
��
��
The isolated letter o consists of the head, the tooth and the circle. Occurring in the initial position, it is proceeded by the connecting line segment. In the middle position it has the shape of a circle, with the head being omitted. In the final position this circle is proceeded by the tail turned to the right and downwards: omolo ������ “grand
gorel-1.p65
56
4/17/02, 8:56 AM
57
manchu script
son.” Occurring at the end of a monosyllabic word, it is written without the tail: fo “a small net attached to a pole (used to fish things out of ice floes).” In the middle and at the end of words the letter o may be doubled: coo ��� “a spade;” cooha ����� “army,” “troops,” “soldier;” doohan ������� “bridge;” guyoo ����� “green jasper.” ��
5) u
��
��
This letter has a shape similar to the letter o, but is distinguished from it by the dot which is placed on the right of the circle. How ever after the letter g when rendering the velar [g] and the aspirated consonant g� as well as after the letter d, the dot does not indicate the vowel u, but these consonants. The dot becomes the means of distinguishing the letter g from the letters k and h (when rendering velars and uvulars and the aspirated consonants k�, g�, x�) on the one hand, and the letter d from the letter t on the other (see the data concerning the Manchu graphemes rendering velar and uvular variants of the consonants /k/, /g/, /h/ in 3.2). 6) å
���
���
���
��
Written separately, this letter consists of the head, the tooth, the circle and the short tail turned downwards. In the initial position the short tail is replaced by the oblique line segment drawn on the left of the connecting line. In the middle of a word it loses both the head and the tooth, and at the end of a word the oblique line segment changes into the short tail turned downwards: akå ����� —particle of nega tion: there is not, there are not, doesn’t exist. This letter has very restricted use in the initial position, occur ring only in words borrowed from Mongolian: ålen ������ “house,” ålet ������� “Oirat,” “Elut,” åren ������ “an emage,” “a doll,” “a Buddhist image,” “a religious image” (Zakharov, 1879:21-2, 55). In the middle and at the end of words this letter is widely used, espe cially with the letters rendering the uvular variants [q], [�], [�] of the consonants /k/, /g/, /h/. Placed on the right of the letter å, the dot serves to indicate the letter g rendering the uvular [�]: gåsa �� � ��� “banner.” In the transliteration of this letter it has been accepted by a number of specialists to write a bar above it: å. In accordance with the Russian tradition in Manchu studies, the same mark, viz. the bar, was used when it was required to render velars [k], [g], [x] in the syllables
gorel-1.p65
57
4/17/02, 8:56 AM
58
part two
gu, hu and the aspirated sounds [k�], [g�], [x�] in the syllables ku, ga, ha; go, ho, ko, but it was written above letters rendering con ka, sonants (Zakharov, 1879:22). In Norman’s lexicon the bar drawn over letters which render these consonants is not in use. The bar was also used to distunguish transliterations of two combinations, yo (io�) ( �� ) and io ( ��� ) (Zakharov, 1879:22).
7) y
��
�
��
��
The letter for the semivowel y has a shape of an oblique line seg ment which ends in the hook turned upwards. This oblique line segment is drawn on the left of the connecting line. The letter y is used to indicate graphemes which render “y-vowels” and occurs mostly in combinations with the letters a, e, o, and u 4: ya
��
��
���
���
ye
���
��
���
����
yo (io�)
��
��
���
���
yu
�� ���
�� ���
��� ���
��� ���
Here are some examples with y-vowels: [ya]—yafan ����� “back yard,” “truck,” yacin ����� “black,” “dark,” yada- ����� “to be poor,” “to be wretched,” yayadambi ���������� —the imperfect finite form of the verb yayada- “to lisp,” “to speak unclearly;” [ye]—yeye ���� “mag got,” “glutinous,” “sticky mud,” “annoying,” “(paternal) grandfather,” yebe ���� “better,” “improved of an illness;” [yo]—yobo ��� “fun,” “play,” “joking,” “a person who likes to play,” yohon ����� “water ditch in the field;” [yu]—yun ��� “rut,” “track,” yuyu- ��� � � � “to starve,” “to go hungry.” The combination of the letters y + u writ ten as ( �� ) (with the dot) is only found when it represents a syl lable of CV type. It is never used after a letter for a consonant where the combination of the letters i + o occurs instead (��� ). In Manchu words, this sequence of letters is pronounced as [io]. Therefore, for instance, the word niyo ���� “swamp,” “marsh,” “slough” was 4 In some other systems of transcriptions, the semivowel [y] is marked by the symbol [j], and the so-called “y-vowels,” are to be considered as “j-vowels.” Most scholars, engaged in Manchu studies, use the symbol [j] to render [ì].
gorel-1.p65
58
4/17/02, 8:56 AM
59
manchu script N Vowels rendered by a grapheme
initial
middle
Writing final
isolated
1.
a
��
��
�
���
2.
e
�
�� ��
��
��
��
(after the letters for the velars [k], [g], [h]; for the consonants k�, g�, h�, b, p)
3.
i
��
��
��
4.
i�
��
���� ���
��
5.
o
��
��
��� � ���1
5.
u
��
��
��� � � �43
6.
å
���
���
��
���
7.
y
�
��
��
��
��
Table 1. Manchu graphemes rendering vowels and the semivowel y
written in old Manchu manuscripts as nio ��� . According to Zakharov, some Manchu words, especially those that are monosyl labic, such as nio ��� (the interrogative sentence particle) and bio ��� (the interrogative form from the verb bi- “to be,” “to exist”) are always pronounced as [nyu] and [byu]. The inventors of the Manchu alphabet probably couldn’t decide which mode for trans literating [yu] would be better, io or yu (with the dot) (Zahkarov, 1879:23-4). 3.2. Manchu Graphemes Rendering Consonants At the beginning of a word
1) n
gorel-1.p65
�
59
In the middle of a word
At the end of a word
�� (before vowels) �� (before consonants)
�� � �
4/17/02, 8:56 AM
60
part two
In the initial position the letter n consists of the head, the tooth and the dot which is placed on the left of the tooth. In the middle of a word it loses the head, so that there remains only the tooth which in the final position is replaced by the tail turned to the right. Occurring before consonants (and after vowels) in the middle or at the end of a word, the letter n loses the dot, so that it can be easily mistaken for the letter a. To avoid confusion there is a simple rule, according to which the letter n usually occurs after the letter ren dering a vowel, standing alone or constituting a syllable with a preceding consonant: anda ������ “a sworn brother,” “bosom friend,” “friend from childhood;” tondo ������ “straight,” “upright,” “loyal,” “fair,” “public.” Nevertheless, even knowing this rule it is easy to be confused when reading words such as algan ����� � � “a net for catching quail,” jurgan ����� � � “line,” “row,” “the right,” “duty,” “ministry,” “board,” amaka ������ “husband’s father” (instead, it is possible to read alaga, juraga, amkan, respectively). The following orthographic rule concerns not only the way of writing the letter n with the combinations ya, ye, yo, yu. Occurring before them, the letter n, as well as some other letters which render consonants, is followed by the letter i: niya, niye, niyo, niyu. But all these syllables were pronounced as [nya], [nye], [nyo], [nyu]. The syllable [nyu] was drastically changed by the Manchus who began to write it with the letters i and o (nio). In accordance with the new writing, they began to read this syllable as [nio]: niohe ������ “wolf;” niolmon ������� “moss;” nioron ������ “rainbow.” The following words include combinations of the letter y with letters rendering vowels a, e, and o, which occur after the letter n: niyahan ������� “puppy,” “whelp,” niyada ������� “late in maturing,” “slow in grow ing,” niyaki ������ “pus,” “nasal and bodily discharge,” aniya ������ “year;”� niyehe ������ � “duck,” niyekse �������� “light,” “thin (of clothing),” niyeniye ��������� “weak willed, lacking initiative,” eniye ������ “mother;” niyo- ���� “to scrape meat from bones,” niyomoàun ����������“drifting ice.” In most cases the letter n is written without a dot in the final position, but there are some cases where the dot is replaced: han ��� “emperor,” “knan;” han ��� “a khan who is subjected to China” (Zakharov, 1879:26). 2) ng
gorel-1.p65
��
60
������ (before a) ����� (before e)
���
4/17/02, 8:56 AM
61
manchu script
The consonant ng [�] may occur only in the middle or at the end of a word. In the first position the letter for this consonant has a shape of the tooth and the short tail turned to the right similar to that of a separately written letter i. In the second position this tail is extended to the left similar to that of a separately written letter e: angga ==C=A) “mouth,” “opening,” “hole,” “pass,” “gate;” anggasi ==C=A=I1 “widow;” enggemu =CC�=�7“saddle;” erdemungge
[email protected]�=�K=CC�=� “virtuous,” “talented,” “moral;” teng ,==E� in: teng seme “hard,” “firm,” “fast,” “solid.” 3) k [q]
�
��
��
�� (before a, o, å) �G (before consonants) ���
g [/] h [:]
0
�D
��
��
It is accepted by the majority of manjurologists that the back con sonants /k/, /g/, /h/ manifest themselves through the uvular [q], [/], [:] or the velar [k], [g], [x] respectively. The alternation of back consonants, velar and uvular—no matter how they are treated by specialists, as different phonemes or allophones which occur in complementary distribution with each other—is directly connected with vowel harmony. Uvulars may be followed only by any of the back vowels a, o, å, and velars may occur with any of the vowels e, i, u. Basically, the letters for the uvulars [q], [/] and [:] have the same shape. Occurring in the initial position, each letter consists of the head formed from the slightly modified tooth and circle. In the middle of a word this head changes into two teeth. To indicate the letter for [/], the dot is used, and the mark for the letter rendering [:] is the circle. Both additional elements, the dot and the circle, are placed on the right of the following letter rendering a vowel: kalka �=��) “shield;” gargan ��=H��=) “branch,” “the earth’s branches,” “branch of a river,” “single,” “odd,” “comrade,” “friend;” hohori 0�D�H1 “the soft cartilage jutting out at the side of the aural cavity.” Occurring between two letters the first of which renders a vowel and the sec ond a consonant, in the middle of a word, the letter for [q] is writ ten by means of two teeth, with two dots on the left. At the end of a word, there remains only one tooth, with two dots on the left. The second tooth is replaced by the tail turned to the right: sakga 5=GJ�)
gorel-1.p65
61
4/25/02, 1:56 PM
62
part two
“old,” “old man;” lak ��� in: lak se “quickly!,” “hurry up!”. Following one of these letters for [q], [�] and [�], the letter rendering the vowel å has the bar drawn above it. 4) k [k]
�
��
��
g [g]
��
���
��
h [x]
��
���
��
The consonants /k/, /g/, /h/ followed by any of the vowels e, i and u are the velars [k], [g], [x] respectively. The letters for all three velars has the same shape which is the specific head like a semicircle which begins and ends nearby the vertical connecting line. The letter for [g] is marked by the dot. The letter for [x] is indicated by the circle. Both diacritic signs, the dot and the circle, are placed on the right of the head. To this semicircle, the letters rendering the vowels e, i, u are added to constitute syllables ke �� , ge � , he ��� , ki �� , gi �� , hi ��� , ku �, gu �, hu �8 . In accordance with Russian tradition, the last three syllables have the bar drawn above conso nants in order to distinguish them from corresponding uvulars. To write syllables [kya], [gya], [hya], [kye], [gye], [hye], [kyo], [gyo], [hyo], the letters rendering consonants are followed by the letter i. Therefore the way of their writing became the following: kiya, giya, hiya, kiye, giye, hiye, kiyo, giyo, hiyo. Here are some examples: kiyafur kifur ��������������� = (onom.) the sound of chewing hard objects, the sound of something smashing, kiyei ����� in kiyei nan hiyang “a type of incense,” kiyokan ������� “a small pointed knife used by an arrow maker,” giyakta ������� “leaves that have fallen from a tree,” giyalu �������“crack,” “fissure,” “defect,” giyen ����� “indigo,” giyolo ������ “the crown of the head,” giyose �������� “a type of meat pastry, a meat-filled dumpling,” hiyatan �������� “railing,” hiyalu ������� “a net carrying bag,” ulhiyen ����� � ��� “gradually,” hiyebele �� � ����� “black-eared kite,” hiyor ������� in hiyor hiyor se- “to be robust,” hiyooàun ��������� “filial.” The syllables [kyu], [gyu], [hyu] are written with the letters i and o, as kio, gio, hio, especially in Chinese words, with the exception of the word giyån ������� which denotes a measure of weight equivalent to thirty catties. But in the native Manchu words, these syllables, similar to the syllable nio are pronounced according
gorel-1.p65
62
4/17/02, 8:56 AM
63
manchu script
to their writing: kiongguhe ������� � � � � “myna of South China (acridotheres cristatellus),” gioro ������ “a name of a Manchurian clan (The Aisin Gioro, a name of the imperial clan).” Occurring in the middle of a word, between two letters the first of which renders a vowel and the second a consonant, the letter for the velar [k] has a shape of the semicircle which does not reach the connecting line. At the end of a word, a head is proceeded by the tail turned to the left. The following words include the letter for the velar [k] in the middle and at the end of a word: keksen ����� “joy,” “gratification,” hekcehun �� � ����� � “falling of the flood water in autumn,” kek �� in: kek se- “to be pleased,” “to be gratified,” “to be refreshed,” kukduri �������� “braggart,” “boaster,” gukdu gakda ������� ������� �“with ups and downs,” “uneven,” “unlevel,” hukàen ������ “falcon’s hood,” “a falcon kept in the house.” 5) k� g� h�
� � �
�� �� ��
�� �� ��
Aspirated k�, g�, h� are only found with the letters rendering vow els a and o, chiefly in foreign words, Chinese or Sanskrit. The let ters for them have a shape similar to that for the velars, but the head is crossed by the stroke. Followed by the letter for the vowel o, they are written in the same way as syllables ku [ku], gu [gu], hu [xu], i.e. with velar pronunciation of the consonant. The only difference is that the letters for the aspirated consonants have a short stroke which crosses the semicircle that is the sign for the head of the letter (Zakharov, 1879:30-1). 6) b
�
��
�
The head of the letter b is formed from the circle divided by the short vertical line segment into two halves. In order to form a tooth indicating the letter a, this circle may be extended over the vertical connecting line: ba �� . To indicate the letter e, the dot is placed on the right of the letter b: � . Occurring in the final position, the graphemes for the syllables ba and be have tails turned to the left, but in the case of the syllable be, the dot is put on the right of the letter b: ��� , �� . In the middle of a word, when the letter b is not followed by a letter for a vowel, the head which is the circle does
gorel-1.p65
63
4/17/02, 8:56 AM
64
part two
not pass over the vertical connecting line: � . In the final position this head is proceeded by the tail turned to the left, but the downstroke is longer than that for the letter o: � (o � ). When followed by the combinations ya, ye, yo and yu [io], the letter b is proceeded by the letter i, the way it is done with the letters for other consonants: biya ���� “moon,” “month,” ubiyada ��������� “disagree able,” biyabiyahån ������������ “pale,” “wan,” biyoran ������� “a cliff of red earth.” The syllable [byu] is written as bio ��� , not biyu. �
7) p
��
��
The way of writing the letter p is similar to that of the letter b. The only difference is that the second half of the circle that forms the head for the letter p, has a special interception. Here are some words containing the letter p: pampu ����� “a thick padded coat;” pekte pakta ������� ���� “perplexed,” “dismayed,” “dumbfounded;” pelehen ��������� “a name for the crane,” pipuri ����� in pipuri ilha “the flower of the loquat tree.” �
8) s
��
��
The shape of the letter s resembles an acute angle formed by two line segments. The lower oblique line segment, directed to the left, is attached by the tooth. This shape is similar for two of the three positions of the letter s, at the beginning and in the middle of a word. Changes only happen with the letters which render vowels when they constitute syllables with the consonant s. In the final position, the tooth is replaced by the almost horizontal tail turned to the right. Similar to other cases, before the combinations ya, ye, yo, yu the letter s is followed by the letter i. The syllable [syu] is written with the letters i and o: sio ��� . Here are some examples with the letter s: sasa ���� “together,” useri ������ “pomegranate,” ashan ������ “side,” “appendage,” “peripheral,” “subordinate,” sisiku ������� “a vessel filled with ashes or sand in which one can insert incense sticks, candles,” “a vase for flowers,” siyal ����� “mister,” “gentleman,” siyangci �������� “chess,” siyoo ����� (< chin. xiao) “saltpeter,” “flute,” siolembi �������� — imperfect finite form of siole- “to embroider,” sioi ���� “foreword,” “preface (to a book),” siyån IEOK) in siyån fu “provincial governor.”
gorel-1.p65
64
4/17/02, 8:56 AM
65
manchu script 9) s�
�
��
�'
The way of writing the letter à is similar to that of the letter s. The only difference is that the short stroke is attached to the upper line segment which is a part of the head. The following words contain the letter à in combination with different vowels: aàs�an =��=) “movement,” “vibration,” “behaviour,” s�asiàame �=IE�=�)� in s�asiàame da- “to blow from the side (of the wind),” “to cut (of the wind),” àeàempe �A�A�F=�� “a wasp,” àårgeku ��EH� “reel,” “spool,” àuàu �K�7 “purple,” “sorghum.” 10) t (before t (before d (before d (before
6
�J
�)
,
�@
�)
6]
�J]
��
,]
�@]
��
a, o, i) e, u, å) a, o, i) e, u, å)
There are two series of letters for the dental stops t and d, and the way of writing each of them depends entirely on the vowel with which they constitute a syllable. Followed by any of the letters for the vowels a, o and i, the letters for the fortis t and lenis d are written in the following way. In the initial position the letters t and d have the shape of the round head. In the middle of a word both letters are formed by the oblique line segment which ends with the circle. The graph eme is written on the left of the vertical connecting line. There is only one difference between the letters for the fortis t and lenis d, that is the dot which marks the letter d. Normally the dot is placed on the right of the next letter which renders a vowel. Followed by one of the letters for the vowels e, u or å, the letters t and d have a different shape. In the initial position, the letters t and d are consti tuted by the round head located to the right of the vertical connecting line. The latter is extended above the letters t and d. In the middle of a word, this grapheme changes its shape into an oblique line segment located to the left of the vertical connecting line. This oblique line segment has the small circle. The letter for the lenis d is distin
gorel-1.p65
65
4/25/02, 12:35 PM
66
part two
guished from the the letter for the fortis t by the dot wrtitten on the right of the following letter which renders a vowel. Occurring after a letter for a vowel and before the following let ter for a consonant, the letter t underwent some changes, that is, the round head was carried from the left onto the connecting line and added with the tooth ( �� ). At the end of a word this tooth is replaced by the tail turned to the right ( �� ). The final letter t can be mistaken for on, but it should be remembered that such shape of t may occur only after a letter which renders a vowel: bithe �������� “book,” “letter,” tuttu ������ “like that,” “thus,” “so,” tubet ����� “Tibet,” “Tibetan,” oirat �������� “Oirat.” According to Mongolian tradition, the letter t, followed by the letter for the back vowel å, is pronounced as [d] inspite of the fact that it has no dot placed on the left: tåmbi �������is read [dumbi]— the imperfect form from the verb du- “to hit,” “to strike;” butån ������� is read [budun] “crock,” “large jar” (Zakharov, 1879:37). The following words contain the letters t and d, both variants of them: tatan ����� “a camp,” “a stopping place,” “territory of a tribe;” dacukan �������� “rather sharp;” dodori �������� “a hat with a wide brim;” dise �� � ��� “draft (of a document or essay);” diyan �� � ��� “palace,” “throne room,” “great hall,” “hostel,” “inn;” ederi tederi �������������� “here and there,” “this way and that;” tiyelin ������� “a short, sharp-pointed arrow used for bird hunting,” tinggin �������� “office,” “bureau,” “section,” latihi ������� “torn piece of a mat,” bandi ������ “learned man,” “pundit,” soti ���� “one name for the parrot,” ududu �������� “several,” “a number of.” �
11) l
�
�
In the initial position, the letter l is written through the head, simi lar to that of the letters a, e, i, o, u, å, n, the tooth and the hook turned upwards and placed on the right of the connecting line. In the middle of a word the head is omitted, and in the final position there re mains only the tail with the hook turned to the right and upwards. Followed by letters which render vowels, the letter l keeps its shape stable. In accordance with the principal rules, all changes happen with letters which represent vowels. The following Manchu words contain the letter l in different po sitions within words: lala ���� “end,” “last,” “final,” “cooked gluti
gorel-1.p65
66
4/17/02, 8:56 AM
67
manchu script
nous millet or rice;” bolori ��H1 “autumn,” “fall;” lulu �K�7 “or dinary,” “common,” “average (of ability);” huluri 0�E�KH1 in huluri malari “careless,” “lax,” “hasty;” liyeliyehun �EOA�EOA�) “dizzy,” “faint,” “delirious;” lioho �E�D� “a white-striped sea fish,” “Siberian salmon;” algin =�CE�) “fame,” “the male otter;” mandal �==J�=� “the mandala (Buddhist).” 12) m
�
�
�
The shape of the letter m is similar to that of the letter l, but the hook is turned downwards. In the final position the letter m consists of the tail turned to the right and the hook turned downwards. The combination of m + yu is written as mio: �E� . In Chinese words it is pronounced as [myu], but in native Manchu words it is pro nounced like [io]: miosihon �E�IED�) “evil,” “false,” “heretical,” “heterodox;” miodori �E�J��H1 —in miodori miodori “walking with dif ficulty (because of a sore back).” The following words contain the letter m in different positions within words: mamari �=�=H1 —PL of mama “grandmother,” “female ancestor,” “old lady,” “pocks,” “rash;” muke �KC=� “water,” “river,” “stream;” namu �=�7 “ocean,” “sea;” miyori �EO�H1 “crust;” miyoocan �EO��?=) “gun,” “musket;” miosiri �EO�IEH1 “smile.” 13) c
+
�?
��
The letter c resembles a right angle one side of which is parallel to the connecting vertical line. Its shape remains invariable at the beginning and in the middle of a word. The letter c never occurs at the end of Manchu words. Here are some instances of words with the letter c: cacari +=?=H1 “a tent;” cahan +=D=) “buttermilk,” “sour milk,” cecike +A?EC=� “a small bird;” cucu caca +K?7 +=?) (onom.) the sound of whisper ing or talking in a low voice; cirku +EH “pillow;” cise +EI)� “veg etable or flower garden;” ucun K?K) “song,” “ballad;” ucuri K?KH1 “time,” “opportunity;” cån +�E) in cån moo “Chinese cedar.” 14) j
�
��
��
In the initial position the letter j is written like the letter y, but the oblique line segment has no hook. In the middle of a word this line
gorel-1.p65
67
4/25/02, 12:42 PM
68
part two
segment turns upwards in a slightly arched way. It seems the letter j never occurs at the end of a word. The following words contain the letter j in different positions within words: ajigen ������� “young,” “small,” “a youth;” jakdan ����� � � “pine;” jeje ���� “father;” jiji jaja ��������� “twittering;” joman ����� “edge,” “seam,” “end;” julen ����� “story,” “tail;” hojo �����“beautiful,” “attractive,” “pleasing;” uju ���� “head,” “first,” “the first;” jiyan ����� in jiyan hooàan “a lip of paper;” jiyanggiyån ������������� “general.” ��
15) r
�
��
In all positions of a word—in the middle or at the end, the letter r has a similar shape. It consists of the head formed by two short line segments crossing each other and attached to the connecting line. In a case when the letter r occurs at the end of a word, this line is extended a little further and ends with the short tail turned to the left. In Manchu there are no native words beginning with the conso nant r. At the beginning of a word the consonant r can only be found in loan words, borrowed chiefly from Sanskrit. The syllables [rya], [rye], [ryo], [ryu] only occur in loan words in Manchu as well. But they are written according to the same method which is used to render combinations of some other conso nants, e.g. n, k, b, s, l, m, c, j, etc. with graphemes ya, ye, yo, yu. The syllable [ryu] is written with the letters i and o, i.e. rio: ��� . Here are some examples of words which contain the letter r: rakca (raks�a) ����� “a man-eating demon,” “ogre,” “Russian;”koro ���� “sorrow,” “regret,” “damage,” “injury,” “offense,” “wound;” urun ����� “daughter-in-law,” “wife;” lor ���� —in lor seme “speaking in cessantly,” “talkative;” berhu �� �8 —term of address used by a wife to her husband’s younger sister, or by an elder sister to her younger. 16) f (before a, e) f (before i, o, u, å)
� �
� ��
�� ��
The letter f begins with the head which is drawn on the right of the connecting line. This head crosses the connecting line from the right to the left, turned into the hook. The letter f keeps this shape to render
gorel-1.p65
68
4/17/02, 8:56 AM
69
manchu script
syllables with the letters for the vowels a and e at the beginning or in the middle of a word. In order to be added with letters for the vowels i, o, u, å this grapheme loses the head and takes the shape of the letter v (w) which is the same as that of f only without the head. That is why the syllables vi (wi), vo (wo), vu (wu), vya (wya), vye (wye), vyo (wyo), vyu (wyu) are absent in Manchu, and it is easier to write syllables fi, fo, fu, få. with the letter v(w) than to do it with the little more complicated letter f. It seems that there are no words ending in either the consonant f or the consonant v (w) in Manchu. The syllable fio 9E� , written with the letters i and o, in some words is pronounced as [fio] but in other words it sounds as [fyu]. The following are some words containing the letter f followed by different vowels: fafun .=MK) “law,” “decree,” “prohibition;” ufa KB) “flour,” “meal;” efen BA) “bread,” “any sort of bread like product made from flour;” feye .AO “nest,” “wound,” “the eye of a needle;” fufun 9KMK) “a saw;” fiyan 9EO=) “rouge,” “luster,” “ap pearance,” fiyen 9EOA) “powder,” “the feathers on the arrow shaft;” fiyoto 9EO�J� —“to break wind,” fiyoose .EO��I “a gourd dipper,” “a ladle;” fioha 9E�D) “pullet,” “young tender chicken.” 9
17) v (w)
�9
��
The shape of the letter v (w) is similar to that of the letter f, but the former has no head in any positions where it occurs, viz. at the beginning or in the middle of a word. It may be followed only by the letters which represent the vowel a and e. The following words include the letter v (w) within the syllables va (wa) and ve (we): wa 9) “odour,” “smell;” wacan 9=?=) “a protec tive covering for the armpits on armor;” wadan 9=J�=) “a cloth wrap ping,” “curtain around a sedan chair,” “cloth of a flag, flag, ban ner;” we 9 “who?;” weren 9AHA) “ripples on water,” “hoop (on a barrel, tub, etc.,),” “a wire circle inside a hat;” gåwa ���E9) “other,” “another;” wesibun 9AIE�) “advancement,” “lifting up;” s�uwe �KM “direct,” “straight,” “totally,” “(with negatives) not at all.”
gorel-1.p65
69
4/25/02, 12:43 PM
70
N
part two
Consonants rendered by a grapheme
initial
1.
n
�
2.
ng
3.
k [q]
4.
g [G]
5.
h [�]
6.
k [k]
7.
g [g]
8.
h [x]
9.
b
10.
p
11.
s
12.
à
13.
t (before a, o, i) t (before e, u, å)
14.
d (before a, o, i)
15.
l
16.
m
17.
c
18.
j
19.
r
20.
f (before a, e)
d (before e, u, å)
f (before i, o, u, å) 21.
v (w)
Writing in different positions middle
�� (before vowels) �� (before consonants) �� ����� (before i, o, u, å) ����� (before e, i) � �� (before a, o, å) �� (before consonants) �� ��� � �� � �� �� ��� �� ��� �� � � � �� � �� � �� � �� � �� � �� ��� (before consonants) �� ��� �� ��� � � � � � �� � �� �� � � � � �� � ��
Table 2. Manchu graphemes rendering consonants
gorel-1.p65
70
4/17/02, 8:56 AM
final
�� � ��� �� �� �� �� �� �� �� �� �� ��
�� �� � � �� �� �� �� �� ��
71
manchu script 3.3. Additional Graphemes
As it was mentioned above, some additional graphemes (ma. tulergi hergen “foreign letters”) were included into Manchu for transcribing foreign sounds, mostly Chinese ones. Partly they were constructed on the basis of elements from which the letters of the Manchu al phabet were formed, partly from ones specially devised for that purpose. The letters for the aspirated back consonants k�, g�, h� which occur in loan words from Chinese have been already described in 3.2. As it was shown, they are slightly differ from the letters for the velar [k], [g], [x]. The other additional graphemes devised by the Manchus specially for transcribing Chinese syllables, are the following ones. It should be noticed that the transliterations of the graphemes for these addittional Chinese consonants and syllables are given according to the Möllendorff’s system. Their transcriptions as they are accepted in the People’s Republic of China according to pinyin system (con temporary Chinese phonetic alphabet), are given in brackets. �
1) ts’ (c)
If the acute angle which is exactly the shape of the letter s crosses with the short vertical line segment parallel to the connecting line, not far off its vertex, we obtain the letter for ts’ (romanized accord ing to Chinese rendering as c). This letter keeps its shape in all positions in a word. It is followed by vowels which change in ac cordance with normal rules of writing. ���
2) ts (ci)
To indicate the syllable, romanized according to Chinese render ing as ci, the special sign “ � ” is added to the grapheme which renders the syllable ts’a (ca). ���
3) sy (si)
In order to write the syllable sy (si), the same sign “ � ” is attached to the grapheme which renders the syllable sa. �
4) dz (z)
To write the letter for the sound dz (romanized according to Chi nese rendering as z), the short vertical line segment is drawn across the vertex of the acute angle of the letter s. As one can notice, there
gorel-1.p65
71
4/17/02, 8:56 AM
72
part two
is only a slight difference in the way of writing the letters for ts’ (c) and dz (z). ��
5) c’y (chi)
To render the Chinese syllable chi (as romanized according to Chi nese rendering tradition) which was absent in Manchu, a special grapheme was devised. To form this grapheme the circle was added to the right of the Manchu grapheme which rendered the syllable ci. ��
6) jy (zhi)
To render the Chinese syllable zhi, as romanized according to Chinese tradition, a new grapheme was invented. It was formed from the Manchu grapheme which rendered the syllable ji by adding the circle on the right of the letter j. 7) z� (r)
�
To render the Chinese sound r, as romanized according to Chinese tradition, a special sign was devised. It consists of a horizontal line drawn above the connecting line. All changes occur with the letters which render the following vowels.
4. Outward Appearance of Manchu Writing Manchu letters are written from top to bottom therefore lines are drawn vertical after the Chinese pattern of writing, but contrary to it, from the left to the right. Both cursive and type initially followed the Mongolian patterns, but later they acquired a slightly more rounded shape. In the middle of the nineteenth century there were three kinds of Manchu script in use: 1) the standard form of the script (ma. ginggulere hergen; ginggule“to write the standard form of the script”), 2) the semicursive script which is similar to the standard form, but letters were written slightly closer to each other (ma. gidara hergen; gidame ara- “to write in the semicursive script”), 3) the cursive script (ma. lasihire hergen; lasihime ara- “to write the cursive script”). This script is chiefly distinguished from the others by tails (ma. uncehen “tail”), which were rounded with flourish. Besides, various additional line segments were written in a more free manner (Zakharov, 1879:48).
gorel-1.p65
72
4/17/02, 8:56 AM
73
manchu script
In all cases—no matter how important the text is— Manchu letters are written in the same size. There are no capital letters in Manchu script. Following the Chinese pattern, words denoting the Emperor, high officials, older relatives or polite forms of address should be indented. The higher the words are above other lines, the more respected are the persons. This etiquette is strictly observed with respect to the ruling dynasty. The title of the ruling emperor is written one word higher above a line, the title of his father is placed two words higher, and that of his grandfather is located higher by three words. Sometimes all honorary titles are written within the same line but with a space of one or two words, according to the importance of a person. The same rule is strictly observed in respect to all words denot ing bureaucratic activities and all sorts of realia connected with them, for instance: hese “imperial order,” “edict;” wasimbu-“to issue ( an order),” “to send down an edict;” wesimbu- “to submit, to present (to the Emperor),” “to report to the throne;” donjibu- “to notify the Emperor;” kesi “favour,” “grace,” “kindness,” etc. All these words should be indented, or can be written within the same line but with a space of one word. The same etiquette was observed in official affairs and diplomatic relations as well as between private persons who were in correspon dence with each other. It was not allowed to carry a word over to the next line. To avoid division of a word, one should calculate the space with the number of lines and words in lines. If blank space remained final tails of letters were written rather longer or markers of cases were written separately from words. But normally all markers of cases, the inter rogative particle ni, the sentence (predicative) particles kai and bi as well as all words with which the speech is ended, should be placed in the same line.
5. Punctuation Marks In Manchu there were two punctuation marks, viz. the full stop �and the colon �. The full stop was drawn across a line and used in the following cases: 1) to separate words which formed the subject NP from that of the VP; 2) to separate different clauses which rendered
gorel-1.p65
73
4/17/02, 8:56 AM
74
part two
the subject and the predicate; 3) to separate one syntactic period from another. The colon was put also across a line and used to mark termina tion of speech, hence at the end of a chapter, an article, a book as well as a letter or a document. A question mark was replaced by special particles which followed the words with which an interroga tive sentence or a syntactic period were ended. To separate different articles and sections of laws, treaties, the large circle was used before each article. This rule was similar to the Chinese and Mongolian ones. The same sign was used when it was required to enumerate objects of any kind. Titles of articles, chapters of books, quotations, especially edicts of emperors and dicta of wise men always required a new paragraph. Those extracts were always written a word higher than all the other lines of the text however long they were.
gorel-1.p65
74
4/17/02, 8:56 AM
75
manchu script
PART THREE
PHONOLOGY
gorel-1.p65
75
4/17/02, 8:56 AM
This page intentionally left blank
77
phonology
PART THREE
PHONOLOGY 1. Vowels and Consonants The reconstruction of the original pronunciation became one of the most important problems for Manchu linguistics. The investigation of the Manchu sound system has not been successful in every respect up to now because the phonation which became known to us in the past century from Chinese and European transcriptions, do not reproduce the real phonetic value of the graphemes. These transcriptions were mastered and made widespread by the Peking (Beijing) Manchus whose pronunciation was influenced to a large degree by the Northern dialect of the Chinese language. 1.1. Vowels At the very beginning of the twentieth century Russian scholars devoted special attention to studying Spoken Manchu (probably, its dialects rather than Manchu itself) in order to reconstruct the phonetic system of classical Manchu (Muromskii, 1906, 1907-8; Grebenschikov, 1912; Shmidt, 1907, 1908). According to them, the Manchu vowel inventory originally included eight phonemes. Four back vowels were in opposition to four front vowels, and together they constituted four pairs: a and ä, o and ö, u and ü, � (like the Russian phoneme /� / orthographically written as ; in the Cyrillic alphabet) and i. By the time the Manchu script was created some of them had been lost and there remained only six vowels: a, o, å (spelled as y� in the Cyrillic alphabet and called “long” or “hard” by Zakharov), e, i, u (spelled as y in the Cyrillic alphabet and called “soft” by Zakharov). According to P. G. von Möllendorff, Manchu writing distinguished six vowels, although, in reality there were eight: a, ä, o, ö, y, i, u, ü. Four guttural vowels a, o, y, u were opposed to four palatal vowels ä, ö, i, ü (Möllendorff, 1892:1). In connection with this some questions arise. The main problem is the characterization of the Manchu vowel e. In accordance with
gorel-1.p65
77
4/17/02, 8:56 AM
78
part three
the Russian script, I.I. Zakharov transcribed /e/ as the Russian grapheme �, basing himself on the similarity of pronounciation of the Manchu /e/ and the Russian [�] spelled as � in Cyrillic alphabet (Zakharov, 1879:52- 3). A. Castren transcribed the phoneme e as ä, P. P. Shmidt designated it as ö. P. G. Von Möllendorff brought the vowel e into correlation with both ä and ö, basing his opinion on the fact that Manchu vowel e corresponds both with ä and ö in Tungusic: ma. ejen “master”—tung. äjän; ma. inenggi “day”—tung. inängi; ma. elge- “to lead”—tung. ölgöja-; ma. edun “wind”—tung. ödyn (Möllendorff, 1892:1). V.I. Tzintzius described the vowel e, common to all Tungus-Manchu languages, as mid-central sometimes slightly rounded (Tzintzius, 1949:78). Gisaburo N. Kyiose believes that the sound value of the Manchu /e/ is the mid-central vowel, that is, shwa [�]. The Manchu /e/ derived from the Jurchen /e/ , the sound value of which was the mid-front [e], i.e. jur. [e] > ma. [�] (Kiyose, 1997:150). The specialists in Sibe still differ in their evaluations concerning the nature of the phoneme e and its variants. In V.V. Radlov’s opinion the Sibe vowel e does not correspond to the Russian /e/ (spelled in the Cyrillic alphabet as e) or [�] (spelled in the Cyrillic alphabet as �): in contrast, it is a palatal sound more comparable with the guttural sound [a] (Radlov, 1888). The vowel e might be put into correlation with the vowels e and �, as they are given by S. Hattori and K. Yamamoto, and with the vowels � and � envisaged in the phonetic interpretation suggested by J. Norman. These phonetic interpretations describe � as a mid-central vowel, and e and � as mid-front vowels, with e rising higher than � according to IPA (Yamamoto, 1969; Norman, 1974). S. Hattori and K. Yamamoto specify the phoneme /e/ and interpret it phonetically as [�]: /�ere/ [��r] “this;” /\el�d�N/ [��ld�n] “light” (ere and elden in Manchu, respectively). The texts recorded by V.V. Radlov suggest that although his material was described in phonetic terms, i.e. on the level of sounds, the vowel /e/ was specified from phonological considerations, i.e. on the level of phonemes. Li Shulan mentions at least two vowels � and � which are absent in Literary Manchu. They originate from the diphthong ai and and the vowels a and o occurring in the initial positions and followed by i: ai > �; aCi > �; oCi > � : ma. aiha [ai�a]—sib. [��] “glazed pottery,” “crockery,” “porcelain;” ma. alin [alin]—sib.[�lin] “mountain;” ma. dobi [dobi]—sib. [d�vi]
gorel-1.p65
78
4/17/02, 8:56 AM
79
phonology
“fox;” ma. omimbi [omimbi]—sib. [�mim] = IMPF of omi- “to drink” (Li Shulan, 1983:303-4). As for the distinct vowels u and å, they are probably variants of one phoneme. One of them, the vowel u, its sound value is [u], occurs mainly with preceding velars: [ku], [gu], [xu], and the other, the vowel å occurs mainly with uvulars: [q� ], [�� ], [�� ]. L. Ligeti believed that the different letters for the phoneme u, being two graphic variants, are only orthographic means to distinguish velars [k], [g] and [x] from uvulars [q], [�] and [�] respectively (Ligeti, 1952). H.C. von der Gabelentz and C. de Harlez indicated them as u and o � respectively (Gabelentz, 1832; Harlez, 1884). E. Haenisch also designated them as u and o � (Haenisch, 1986). From the very beginning of studying Manchu, the Russian linguists used the horizontal bar over the vowel å (spelled as � � in the Cyrillic alphabet) not to designate a long vowel but, contrary to the existing linguistic tradition, to mark a distinct vowel, different from the vowels o and u (spelled as y in the Cyrillic alphabet). Zakharov claims that the “hard” vowel å represents two sounds. One of them is used to render the Mongolian sound which occurs in the initial position of several loan words from Mongolian, for example: ålen ������ “house,” ålet ������� “Oirat,” “Elut,” åren ������ “an image,” “a doll,” “a Buddhist image,” “a religious image.” The other is used to indicate preceding uvulars [q], [�], [�]. The vowel å also occurs, but very seldom, after other consonants m, s, à, t: månggu ������� “bird’s nest (the edible nest of a type of swallow),” såna ����� “pack,” “a leather leash for a dog,” àårgeku �������� “reel,” “spool,” tå- (då-) “to hit,” “to strike,” ( tåmbi ������ = IMPF of tå-) butån (budån) �� � ��� “crock,” “large jar.” According to Zakharov, in the last two positions, viz. after uvulars and some other consonants, the “hard” vowel å is pronounced similar to the “soft” vowel u (Zakharov, 1879:21-2, 55). Thus, in Zakharov’s view, the hard vowel å is chiefly used to indicate uvulars [q], [�], [�], for instance: akå ����� “there is not/ there are not,” “doesn’t exist;” gåsa ������ “banner;” kåthåmbi ����������� = IMPF of kåthå- “to mix,” “to mix up,” “to stir,” “to stir up,” “to confuse.” One can assume that in such a position it should be pronounced slightly more open and back, like [�], than its higher counterpart, the soft [u]. Based on the similarity of pronounciation of the Manchu and Mongol vowels, the Chinese scholar Qinggetaier suggested the following description of three round back vowels o, u, and å (o � or
gorel-1.p65
79
4/17/02, 8:56 AM
80
part three
� � in some other interpretations) in Manchu. The sound value of the vowel o is [�]; the high round vowel u is similar to the Mongolian [�u]. The most questionable vowel å (o,� � � ) is a round centre open sound [o]. It is used to render two sounds which occur in complementary distribution. One of them regularly occurs after the consonants k, �, x (clearly, these symbols indicate the uvular sounds [q], [�], [�] respectively – L.G.), and its sound value is [�]: k�, ��, x� ([q�], [��], [��] – L.G.). The other occurs in the initial position of Mongolian borrowings, for example: mo. ögeled—ma. ågeled; mo. öbesüben—ma. åbesuben; mo. öngge—ma. ångge. In this position its sound value is soft [o] spelled as ö in Mongolian. According to him, Manchu vowel inventory should be properly represented as follows: a, �, i, �, u, o (Qinggetaier, 1985:398-402). V.V. Radlov mentioned one more sound transcribed by the grapheme ω and described this vowel as the “so called close o as in the German word Monat” (Radlov, 1888:18). Presumably he used the symbol ωto mark the vowel in question which was transliterated by the symbol å (or o� in some other interpretations). The grapheme ω probably corresponds to the sound attuned somewhere between [u] an [� ]; in L.V. Scherba’s General Phonetics the sound is transcribed by [o] (V. Kotvich transcribed it as [�]). It is possibly the same sound [�] mentioned above. In addition to the sounds included in the S. Hattori and K. Yamamoto system, J. Norman postulates the labialised high front sound [ü]. The fact that this sound is an allophone rather than a phoneme becomes clear from the phonetic materials given by S. Hattori and K.Yamamoto, whose phonological description comprises the diphthong /iu/ realized phonetically as [�]. From the aforecited survey it becomes clear that the monoph thongs listed by different authors are more or less comparable in quantity, though in quality they are different. Diphthongs, as they are treated by different linguists, are much less comparable in quantity than monophthongs. According to Russian tradition, in Literary Manchu it is common to distinguish only four descending diphthongs (with the narrow final element): ai (aisi “benefit,” “profit,” aisin “gold,” bai “plain,” “simple,” “free,” “unemployed,” “only,” etc., dailan “a military campaign,” “a punitive expedition”), ei (eigen “husband,” eifu “tomb,” “grave,” deijiku “firewood,” meihe “snake,” “the sixth of the earth’s branches”), oi (oilori “on the surface,” “on the outside,” boihon “earth,” “soil,” “dirt,”
gorel-1.p65
80
4/17/02, 8:56 AM
81
phonology
koiman “tricky,” “cunning,” “deceitful,” oilon “surface”), ui (uihe “horn,” uilen “service,” “attendance,” guifun “a ring for the finger,” guilehe “apricot,” sui “crime,” “guilt,” suila- “to work hard,” “to be exhausted”) (Pashkov, 1963:16). This is unlikely to reflect the situation adequately. It is possible to specify several more diphthongs, four of them being rising ones (with the wide final elements): ia (giyalan “space between,” “interval,” “interstice,”giyamun “stage,” “relay station,” “military post station,” giyahån “accipitrine birds: hawks, falcons, etc.,” hiyase “box,” “case,” miyali- “to measure”), ie (fiyen “powder,” “the feathers on the arrow shaft,” hiyese “scorpion,” miyehusu “bean-curd skin,” niyecen “a patch,” “a small piece of cloth”), io (niohon “dark green,” fiyoose “a gourd dipper,” “a ladle”). That these combinations are diphthongs becomes clear from the phonetic transcriptions given by S. Hattori and K. Yamamoto in “Classified Dictionary of Spoken Manchu” where most of these words can be found: giyalan /gial�N/ [���al�n] “room space,” “compartment of a house,” giyamun /giam�N/ [���am�n] “stage,” “cheking station,” giyahu�n /giahuN/ [���a�ωn] “hawk,” miyali- /mial�m�, mialim�/ [m��al�m, m��alim] “to measure;” fioose /fios�/ [f ��os] “gourd ladle,” niohon /niooh�N, nioohuN/ [���o��ön] “dark green”1. The descending diphthong eo also occurs in classical Manchu: deo “younger brother,” “younger (of males),” geo “a mare,” jeo “department (a political subdivision),” leole- “to discuss,” “to talk over,” leose “building,” “multistoried building,” “tower”, àeole- “to embroider,” “to collect.” It seems that the diphthong eo in classical Manchu corresponds to the diphthong [�u] in most words in the variety of Spoken Sibe recorded by K. Yamamoto and S. Hattory: geo /ge�u/ [�� u] “a mare,” leole- /le�ul�m�/ [l�ul�m] “to discuss,” “to debate,” “to argue,” leose /le�us�/ [l�us] “two-storied building,” seole� /s �e�ul�m�/ [��ul�m] “to embroider.” There are few words where the diphthong eo corresponds to the long vowel u in this variety of Sibe: deo /duu/ [du�] “younger brother”2. At least two rising diphthongs, viz. ua and ue, appeared as the result of dropping the consonant w in the
1 In “A Classified Dictionary of Spoken Manchu” I found just a few words with the diphthong ie [���], for example: niyengniyeri /ni�ia��ni�iari/ [�(��)��������] “spring.” 2 K.Yamamoto used the symbol s� [�] for à. He presented verb stems in the form of the imperfect converb (-me), for example: da- /dam�/ [dam] “to catch fire;” te- /tem�/ [t�m] “to sit,” “to ride on,” “to dwell.”
gorel-1.p65
81
4/17/02, 8:56 AM
82
part three
intervocalic position: juwan > juan “three;” juwe > jue “two;” kåwaran > kåaran “camp (military),” “market place,” “workshop,” “yard;” suwe > sue “you;” tuwa > tua “fire.” Möllendorff postulated four rising diphthongs: ia, ie, io, iu. According to him, the following six rising diphthongs: ua, åa, ue, åe, uo, åo appear when the consonant w is dropped in the intervocalic position. He also pointed out seven descending diphthongs: ai, ei, oi, ui, åi, ao, eo. The combinations ii and oo were referred to diphthongs by him as well. He also specifed the following triphthongs: ioa, ioo, io(w)an, io(w)en, ioi, i(y)ao (Möllendorff, 1892:3). Radlov delimited ten diphthongs, with seven of them being descending ones: ai, ei, oi, ui, ωi, au (au�), eu (eu�) and three of them being rising ones: oa (o�a), ua (ua� ), ue (u�e) (Lebedeva & Gorelova, 1994)3. According to him, the combinations ija ��� , ije ��� , ijo ��� , iju ��� were used to designate palatalization of preceding consonants (in his transcription the symbol j was used to render the phoneme y – L.G.). Hattori and Yamamoto describe only four rising diphthongs on the phonological level; in pronunciation they are realized as either front diphthongs or front monophthongs related as position allophones: /ia/- [� ~ ia], /ie/- [�], /io/- [� ~ io], /iu/- [�]. It is unclear why the Japanese linguists postulate the diphthongs /ie/ and /iu/ which according to their own material are realized as monophthongs only. One can postulate that this is done with respect to systemic considerations. A comparison made between the number of diphthongs described by S. Hattory and K. Yamamoto and that envisaged by V.V. Radlov and J. Norman suggests that the Japanese linguists did not take into consideration all the diphthongs occurring in their version of Sibe. Thus, for example, the lexical material recorded by them gives us the diphthong ai [aι] (aisin /’a’is�iN/ [?aι�in] “gold;” baita /ba’it�� [baιt] “affair,” “task,” daifu /da�ifuu, da�ifu/ [daιfu�, daιf�] “medical doctor,” faidan /fa�id�N/ [faid�n] “line,” “row,” “file”), the diphthong oi [�ι] (boigon /bo’ih�N, /boihuN/ [b�ι��� n, b�ι�ω�� “independent household,” doigonde /do�ih�nd�/ [d�ι��� nd] “in advance,” “beforehand,” goida/�o�idam�/ [��ιdam] “to take a long time”), the diphthongei [�ι� 3
In Radlov’s “A General Linguistic Alphabet Compiled on the Basis of the Russian Letters” the vowel u is rendered by the symbol y, in accordance with its spelling in Cyrillic alphabet.
gorel-1.p65
82
4/17/02, 8:56 AM
83
phonology
eici /�e�ici/ [��ιJ�] “either … or,” eiten /�e�it�N] [��ιt�n] “each and every,” “all,” beide- /be�id�m�/ [b�ι@�m] “to examine,” “to inspect,” meihe /me�ix�/ [m�ιx] “a snake,” teisu /te�isu/ [t�ιI�] “obligation,” weile /ve’il/ [v�ιl] “crime,” “guilt”), the diphthong ui [uι] (cisui in ini cisui /cisu�i/ [t�ιsuι] “by oneself,” “of one’s own accord,” guise /gu�is�/ [�uis] “food box,” “chest,” suila- /su’ilam�/ [suιlam] “to struggle,” “to work hard,” “to be exhausted,” sui /su’i/ [suι] “Heaven-sent punishment”) and the diphthongua [u�a] (juan /juaN/ [d�u�an] “ten”). J. Norman specifies ten diphthongs, six of them being descending ones: ai, �i, oi, ui, au, �u and five being rising ones: ia, i� , io, ua, ü� (Norman, 1974:163). It is common to distunguish the geminate vowels in Literary Manchu: aa, oo, ee, ii, uu. They are normally transcribed as a, e, i, u respectively, although the geminate oo is transcribed as such more regularly. Old lexicons and dictionaries give /oo/ as the only gemi nate vowel which had been fixed in written form, by the doubled letter oo. But in some scholars’ opinion, ii and oo were considered to be diphthongs, not geminate vowels (see, for example, Möllendorff, 1892:3). Recent studies which have specially examined the vocalic system of the spoken language of the important dialect—Sibe, specify the following geminate vowels: /aa/ [a�], /oo/ [��] /ee/ [��], /ii/ [i�], /uu/ [u�]. Examples of these geminates, adduced above, were extracted from “A Classified Dictionary of Spoken Manchu” by K. Yamamoto: ba /baa/ [ba�] “place,” “locality,” “spot,” buda /b�daa/ [b�da�] “meal,” “surved food,” “cooked cereals,” da /daa/ [da�] “root,” “base,” fa /faa/ [fa�] “window,” ja /jaa/ [d�a�] “easy,” “rea sonable,” “cheap,” jiha /jihaa/ [d�ι���] “money,” “bronze coin,” biya /biaa/ [b��a�] “moon;” boo /boo/ [b��] “dwelling,” “house,” “room,” moo /moo/ [m��] “wood,” “timber,” “lumber,” “log,” solho /sol�hoo/ [s�l���] “Korea,” “Korean,” s �olo / s �oloo/ [�����] “leisure,” too- /toom �/ [t�� ] “to curse,” “swear at;” bithe /bit�xee [bitx�� ] “book,” “document,” fe /fee/ [f��] “old,” “aged,” ice /�icee/ [�it���] “new,” ne /nee/ [n��] “recent,” “of late,” se /see/ [s��] “age,” te /tee/ [t��] “presence,” “now;” bi /bii/ “to exist,” “there is/are,” ci /cii/ [t�ι�] “lacquer,” “lac,” fi /fii/ [fi�] “writing brush;” baju /bajuu/ [bad�u�] “distillation dregs,” handu /haNduu/ [�andu�] “paddyfield,” ku /kuu/ [ku�] “treasure storehouse,” langgu /la�uu/ [la�u�] “pump kin,” su /suu/ [su�] “introduction,” “preface,” “a writing.” All geminates usually occur in word final position. As for /oo/, it may occur in the penultimate syllable: doohan “bridge” (doo
gorel-1.p65
83
4/17/02, 8:56 AM
84
part three Monophthongs front central back
high i �* a
mid low i (Ê)**
�* a i e i ü �
* ** ***
� a � a
Diphthongs
u � o
ai ia ao ua
ei oi ie io eo ue
u
ai au oa ue
ei oi ui ωi (au�) eu (eu� ) (o� a) ua (u� a) (u� e)
ω***
o
ui Literary Manchu
according to Radlov
u o
ia→ [� ~ ia], ie→ [�] according to io→ [� ~ io], iu→ [�] Hattori & Yamamoto
u o
ai au ia ua
�i
�u
oi ui
according to Norman
i� io ü�
� is written as e Ê corresponds to iye in Radlov’s view (Radlov, 1888:37) ω is presumably used to indicate [�]
Table 3. Vocalic inventory (in Literary Manchu and Sibe)
“to cross a river”), doose “a Taoist priest,”doosi “greedy,” “covetous,” hooàan “paper,” kooli “rule,” “norm,” “law,” “custom,” “habit,” “document,” “method,” oori “semen,” “essence,” “spirit,” “energy, cooha “army,” “troops,” “soldier,” “military,” àooge “hard feather at the tips of the wings,” toose “power,” “authority,” “right.” The combination oo in classical Manchu corresponds to [au], [aω], [u�], [�], [��] in the variety of spoken Sibe, recorded by Yamamoto: coo /cuu/ [t�u�] “shovel,” “spade,” doo- /da�um�/ [daum] “to cross,” “to cross over,” “to ford,” doola- /dol�m�, dolum�/ [d�l�m, d�lum] “to pour,” doose /da�us�� [daωs] “Taoist monk,” doosi /da�us�i/ [dau�] “rapacious,” hoosan � /hos� iN, ha�us� aN/ [���ιn, �aω�an] “paper offerings used in ancestry worship ritual,” kooli /qa�uli, qoli/[qaωlj, q�lj] “precedent,” “procedure,” oori /�a�uri/ [� =ωrj] “vigor,” “vitality,” too- /toom�/ [t�� m] “to curse,” “to swear at,” toose /ta�us�/ [taus] “authority.” In Written Manchu the combination oo
gorel-1.p65
84
4/17/02, 8:56 AM
85
phonology
(double o) could be realized either as the geminate vowel [oo] or as the diphthong [ao]. J. Norman points out the existence of geminate vowels occurring only in word final position; in open monosyllables they are obligatory in this position. He transcribes such vowels by digraphs /aa/, /oo/, /ee/, /ii/, /uu/. According to him, the second vowel of the pair is usually realized as a prolongation of the preceding one, but in the case of /ii/ and /uu/, the second vowel is realized as a lower and centralized off-glide: [iÌ �] ~ [ie�] (Norman, 1974:163). Norman also gives some examples of the geminates in the Glossary, attached to his “A Sketch of Sibe Morphology:” asitaa “young man,” budaa “food,” “cooked cereal,” jihaa “money,” sunjaa “five,” f�taa “rope,” am�� “father,” bitx�� “book,” àulx�� “apple,” gu�juu “princess,” handuu “rice plant,” mukuu “water,” susuu “kaoliang,” tondoo “straight,” “frank” (Norman, 1974:173-4). V.V. Radlov distinguished no long vowels in his recordings. Table 3 represents the Manchu vowel inventory in comparison with the Sibe vocalic system as interpreted by the authors mentioned. Place of articulation
Labial
Mode of articulation
Obstruens voiceless voiced Fricatives
Affricates
Bilabial
Labio dental
p b
voiceless voiced
front
dorsal
t d f
k g
s, à
v
back
h j*
voiceless voiced
c� (c) c�z� (cz)
Nasals
m
n
Laterals
l
Flapped
r
�
* [j] corresponds to [y] in other systems where [j] instead is used to indicate [z]. � Table 4. Consonants in Literary Manchu
gorel-1.p65
85
4/17/02, 8:56 AM
86
part three
1.2. Consonants The consonantal inventory of Literary Manchu and its dialects has been described by most researchers somewhat inconsistently: some of the consonants were defined phonologically and others were described in terms of their sound realizations. According to Russian tradition in Manchu studies, the back consonants /k/, /g/, /h/ are realized as the velar sounds [k], [g], [x] or the uvular sounds [q], [�], [�]. The velars and uvulars are considered to be allophones occurring in complementary distributions to each other. The velars [k], [g], [x] are immediately followed by the vowels e, i and u whereas the uvulars [q], [�], [�] are immediately followed by the vowels a, o and å. The vowel å is used to indicate two sounds, [�] after uvulars and “soft” [o] in Mongolian borrowings (spelled as ö in Mongolian). In the Sibe dialect of Manchu, the velars and the uvulars are considered to be distinct phonemes and listed by Norman as /k/, /g/, /x/ and /q/, /� /, /h/ respectively (Norman, 1974:162). The uvulars are specified by Hattori and Yamamoto as the post-velar /q/, /�/, /h/ (Yamamoto, 1969:134, 16-7). In Manchu, Kiyose also specifies the velars [k], [g], [h] and uvulars which he treats as the post-velars [q], [γ], [�] ([γ] corresponds to �� ] and [h] to [x] in IPA). Phonologically, the Russian classification is more justified and its approach to the phonologic consonantal system is more consistent since this classification does not incorporate allophones. Zakharov specified the aspirated consonants k0, g0, h0 which occur only with the vowels a and o: k0amduri “the name of a constellation,” k0apsi “a monk’s habit,” k�o “section (of an organization),” k0ose “a figured textile woven from gold and silk threads on a gauze background,” g0an, g0odarg0a. He noted that the aspirated consonants could be found only in foreign words, primarily, in Chinese and Sanscrit borrowings (Zakharov, 1879:30, 57). Möllendorff singled out the letters k‘, g‘, h‘ especially designed to render the aspirated consonants which occur in some Chinese syllables (Möllendorff, 1892:vi.). The most recent Russian classification of the literary Manchu consonants makes no mention of the aspirated consonants; neither does it contain the voiced back spirant [5] and voiceless dorsal spirant [�] (in Radlov’s transcription; [5] is also indicated as [γ] in other systems). Following the existing tradition, this system makes no mention of the laryngeal [�] which is postulated by the Japanese linguists.
gorel-1.p65
86
4/17/02, 8:56 AM
87
phonology
Table 4 represents the literary Manchu consonantal inventory as interpreted by most Russian scholars. The table is given in the form in which it is found in the book “The Manchu Language” by B.K. Pashkov (Pashkov, 1963:17). In the Sibe dialect of Manchu, Radlov pointed out the presence of the palatalized consonants which have not been specified by the other authors. It can be explained in two ways: 1) either Radlov’s observations are correct and such palatalized consonants existed in the dialect he recorded or 2) he used the palatalized consonants to denote i-diphthongs (linguistic materials concerning diphthongs see in 1.1). One can see that /ia/ has been preserved in the dialect described by Hattori and Yamamoto: wacihiyame /vacixiam� / [vat�xiam] “completely,” “wholly;” biya /biaa/ [bia'] “moon,” “month” (in Radlov’s transcription, resp., vacx�ame and ba� ; for the word “moon,” ”month” he also gives another variant—bia). Radlov specified two aspirated consonants t0 and p0 and used the mark “�“ instead of “� ” to designate aspiration. According to him, the aspirated t� occurs in Manchu words: e.g. tere “this,” but t� ere “living;” the aspirated p� occurs only in Chinese words. Radlov singled out the non-sonorous simple spirant (voiced), back (uvulapalatal) which is designated by the symbol [5] (in other systems it is indicated by [γ]). Judging from Radlov’s texts, he used the symbol [5] to designate a sound which correlates with the sound �D� as a voiced allophone:bucehe, buce5e “died;” oho, o5o “became;” genehei, gene5ei “walked for a long time;” jafaha, jafa5a “caught;” sabuha, sabu5a “learned.” It is difficult to ascertain, what sound is transcribed by the symbol [�]; though Radlov classifies this sound as a non-sonorous simple spirant, voiceless, dorsal, he also included in his system its palatalized variant. The grapheme l stands for the middle [l] as in German; it is never pronounced as the Russian ' [Ñ] in ' a [Ña] (Radlov, 1888:37, 39, 73, 74; Gorelova, 1986:307-10; Lebedeva & Gorelova, 1994:24). Norman mentioned that the fortis stops are aspirated before vowels but unaspirated when they occur before other consonants. Norman believes that voicing does not function as a distinctive feature in Sibe; though the lenis stops are voiceless in initial position and before � #, between vowels and sonorants they are voiced. The velar and uvular frivatives x and h are voiced when they occur between vowels and sonorants but they are voiceless before #. Subsequently, according to him, all consonants except the alveolar stops and the
gorel-1.p65
87
4/17/02, 8:56 AM
88
part three
uvulars are palatalized before high front vowels. As for consonant clusters, they occur only internally in words; clusters of obstruents tend to be either totally voiced or totally voiceless (Norman, 1974:162). Norman seems to consider l and r to be allophones of one and the same phoneme, however, this is not confirmed by other authors’ material. They are different phonemes in Literary Manchu too. The symbol h is used to denote (contrary to tradition) a fricativeuvular sound (according to IPA it ought to be indicated by [�] ). The symbol g denotes a nasal uvular sound correlating with the velar [�]; it is unlikely that Norman is correct in postulating this sound (Norman, 1974:162; Gorelova, 1986:308; Lebedeva & Gorelova, 1994:25).
fortis stops lenis stops nasals fricatives semivowels and liquids
labial
alveolars
alveo-palatals
p b m f
t d n s
(à)
v
l/r
y
c j
velars
uvulars
k g � x
q � g h
Table 5. Consonants in Sibe according to J. Norman
There is only a slight difference in the consonantal inventory of Sibe proposed by Norman, on the one hand, and by Hattori &Yamamoto, on the other. The differences are as follows: the Japanese linguists treat /l/ and /r/ as different phonemes; they do not specify the uvular nasal /g�; yet they include as an independent phoneme the voiceless glottal catch (laryngeal) which, in their opinion, precedes all vowels’ initials. The symbol /’/ is used by them to denote, first, the sound [�] and, second, the syllable boundary: alin /�aliN/ [�alin] “mountain,” “hill;” efen /’ef�N/ [��f�n] “Manchu bread;” ome /�om�/ [�� m] “lake,” “pond;” utuku /’utuku/ [�utwkw] “clothes;” baitangga /ba�it��� / [bait��] “necessary;” huwaitame /ha’it�m�/ [�aιt�m] “to tie,” “to tie up;” kaica- /qa�cim�� [qaιt�ιm] “to cry out,” “to yell;” teifun /te�ifuN/ [t�ιfun] “walking stick,” “staff.” Postulation of the laryngal is motivated probably by the principles of phonetic analysis followed by the Japanese linguists since they are the only ones to postulate the laryngeal.
gorel-1.p65
88
4/17/02, 8:56 AM
89
phonology 2. Vowel Harmony
All Tungusic languages are characterized by such a phonetic phe nomenon as vowel harmony (synharmonism). The mechanism of vowel harmony is understood as follows. All vowels of the language are divided into two mutually exclusive sets (groups) on the basis of a certain distinctive phonetic parameter. A number of scholars believe that there are languages where a third set of vowels, different from the previous two, can be distinguished. The latter contains the socalled neutral vowels (it may be just one vowel). Word stems normally contain vowels belonging to one of the two sets, but may include vowels (or a vowel) of the third group. Vowels in the suffixes have to be adapted to the vowels of the word stem, i.e. they normally belong to the same set of vowels, as the word stem has, or they may contain the neutral vowels. The nature of vowel harmony in many languages and dialects is still unclear and requires a further investigation. The question of the nature of vowel harmony in a language (or a dialect) cannot be successfully solved without having a sufficient experimental acoustic data from it. Unfortunately, only a few Tungusic languages (or dialects) have been sufficiently investi gated experimentally, therefore not many experimental acoustic data is available from them. Generalizing a number of existing interpretations of vowel harmony in the Tungusic languages, E. de Boer came to the conclusion that the traditional way to construe the phenomenon was to consider it as the opposition of front vowels versus back vowels, that is vowel harmony based on velarity. Later on, another idea came into being according to which scholars began to describe vowel harmony as the opposition of closed vowels versus open vowels, that is harmony based on the tongue height. But the vowel harmony in Proto-Tungusic continued to be regarded as based on the front-back distinction (Boer, 1996:121). The Tungusic vowel harmony has been investigated mainly by Russian scholars who used to describe it in terms of “soft” (closed/ front) and “hard” (open/back) vowels. But even using the same terminology and describing this particular phenomenon in the same language, scholars sometimes understood it differently. Quite a few descriptions were suggested for the Evenki language, the most widespread Tungusic language spoken in Siberia. Evenki has many features common to all northern Tungusic languages and therefore
gorel-1.p65
89
4/17/02, 8:56 AM
90
part three
it can be regarded as a typical representative of them. Accurate analysis of existing descriptions of vowel harmony in the literary Evenki language and its dialects was given in the aforementioned article of Elisabeth de Boer. She presented investigations that have been published in Russia and abroad (Boer, 1996:121-133). The literary Evenki language was originally based on the Nepa dialect which was later changed to the Poligus subdialect of the Podkamennaja Tunguska dialect. Those Russian scholars who studied the latter form of literary Evenki believed that the vowel harmony in the literary Evenki language was strongly reduced. According to them, the only soft vowel � alternates with the hard vowel a in suffixes in accordance with the requirement of the word stem. The vowels a and o contrast only in the initial syllable. In other positions their occurrences are determined by labial harmony. At present, in the literary Evenki language, only two neutral vowels ! and � exist (as spelled in the Cyrillic alphabet), and they can harmonize with vowels of the two different sets. In the past, there were “soft” and “hard” variants of each of the vowels ! and �. However, in the course of time, the difference between these sounds, as contrasting phonemes, has disappeared. Any differences in pronunciation—if they exist at all—can be considered allophonic. However, certain word stems with the so-called neutral vowels take “soft” vowels in suffixes, while others require “hard” vowels in suffixes. It means that the word stem continues to be assigned to the original harmonic pattern. This point of view was strongly maintained by O.A. Konstantinova, who wrote the “Grammar of the Evenki language.” She claimed that “soft” and “hard” variants of each of the vowels ! and � can only be found in several eastern Evenki dialects. (Konstantinova, 1964). E.P. Lebedeva adhered to a similar idea regarding Evenki vowel harmony. This theory was suggested in her textbooks written for students of tertiary institutions (Konstantinova & Lebedeva, 1953; Konstantinova, Lebedeva & Monakhova, 1985). G.M. Vasilevich, who was one of the first scholars to study Evenki dialects, distinguished “soft” phonemes �, �� and �, �� and “hard” phonemes �, �� and �, �� both in literary Evenki and its dialects. In her earliest publications, she used different symbols to indicate “soft” (� �¯, K K¯) and “hard” (� Ê, � å) vowels (Vasilevich, 1934, 1936)4. In her later publications she adopted a different approach, using 4
gorel-1.p65
The bar over a symbol is used here to indicate a long vowel.
90
4/17/02, 8:56 AM
91
phonology
only one symbol ! to render both sounds, the high front ! and the high central !. Similarly she used only one symbol � to indicate two sounds, the high central � and the high back �. She justified this by practical reasons stemming from the fact that in 1937 the Latin alphabet was replaced by the Cyrillic to write literary Evenki (Vasi levich, 1948:21-2, 1958:654). In fact, she believed that there were no neutral vowels either in literary Evenki or in its dialects. If the word stem contains the “soft” vowels �, �� and �, �� it may be augmented only with suffixes which contain “soft” vowels. If the word stem includes the “hard” vowels � �� and �, �� it may be followed only by suffixes which have “hard” vowels. As for the nature of vowel harmony in the Evenki language, G.M. Vasilevich and her followers considered it as the opposition between “relatively” front vowels and “relatively” back vowels, that is a category based on velarity. Following the theory presented by Joshua Ard, Boer also suggests that vowel harmony in Evenki is based on relative tongue height which, in turn, depends on the position of the tongue root (Boer, 1996:121-33). Ard’s theory was developed as a result of analysis of several interpretations of the Tungusic vowel harmony. The linguistic data was taken from the book “Jazyki narodov SSSR” [Languages of the peoples of the USSR]. According to Ard, the original ProtoTungusic vowel harmony was based on the opposition of advanced tongue root (+ATR) versus unadvanced tongue root (-ATR) vowels. The original pattern of vowel harmony, in Ard’s opinion, was fully preserved in Lamut (Even) as described by K.A. Novikova. Novikova’s field phonetic material was investigated by means of x-ray, and therefore one could see the position of the tongue root in the articulation of the vowels. Up to date, in Lamut, we find the “soft” (high) vowels i and u, which harmonize with the vowel � in added suffixes, and the “hard” (low) vowels � and �, which require the vowel a in added suffixes (Novikova, 1960:52). The vowel harmony in Lamut (Even) can be presented as the opposition of the “soft” vowels to the “hard” ones: soft: i, i: �, �: hard: �,�: a, a:
u, u: �,�:
o, o: �,�:
��e: ��a:
reduced vowel �� reduced vowel a�.
Later on, in most Tungusic languages, in Ard’s opinion, the opposition shifted to a pattern based on relative height. In the original pattern of vowel harmony, the relative tongue height difference was observed, but only as a surplus effect of the tongue root position.
gorel-1.p65
91
4/17/02, 8:56 AM
92
part three
Gradually, this redundant feature had become the main phonetic parameter on which the vowel harmony was based. In Ard’s view, the position of the tongue root is neutral (-RTR) for relatively high vowels, while in the case of low vowels, the tongue root is retracted (+RTR). Normally, -RTR vowels are somewhat more front than their +RTR counterparts. This is especially true for back vowels (Ard, 1981:23-43). Boer maintains that at the time when G.M. Vasilevich described the Evenki vowel harmony, the linguists could not perceive more than three levels of vowel height. The smaller differences in tongue height, as in the case between -RTR vowels and their +RTR counterparts, were interpreted by them as front-back distinctions. They also had no idea that the category of harmony could be based on a single distinctive phonetic parameter, namely the position of the tongue root. It is precisely the tongue root distinction, that may explain why it was so difficult to perceive by ear the difference between “soft” and “hard” variants of each of the vowels i and u. These difficulties were mentioned not only by G.M. Vasilevich, but also by H.Castren who was the first to produce a grammar of the Evenki language5. Boer’s suggestion about the nature of the Evenki vowel harmony, as based on the position of the tongue root, seems to be corroborated by some recent descriptions of vowel harmony in several Tungusic languages and dialects. Firstly, Jan-Olof Svantesson, who investigated the Solon language spoken in Inner Mongolia, came to the conclusion that the Solon vowel harmony was based on the tongue root position. Secondly, two interpretations of the Orochon vowel harmony were recently presented by Chinese scholars (Orochon is considered a dialect of Evenki in China). In the dialect of the village of Gankui, on the banks of the Gan river, Z. Hu discovered two mutually exclusive sets of vowels: “soft” vowels �, �� �, �� �, �� �, �� ��� “hard” vowels �, �� �, �� �, ��� �, �� ���*. * The symbols �� and �� are used here to indicate the prepalatalized long vowels because the latter palatalize preceding consonants or appear after j (Boer, 1996:127).
5 A grammar of the Evenki language, written by H.Castren, was based on the Urul’ga and Man’kova dialects.
gorel-1.p65
92
4/17/02, 8:56 AM
93
phonology
In Boer’s opinion, the difference between “soft” and “hard” vowels in the system as it was described by Hu, is that of relative tongue height (Hu, 1986). A second version of the Orochon vowel harmony (a dialect spoken in the village of Xunke along the Kurbin river) also shows that all vowels can be divided into two mutually exclusive sets on the basis of a very specific parameter, namely “tenseness” versus “laxness.” This parameter can finally be interpreted as based on the differences in tongue position. The “lax” vowels � , o, u and � e correspond to the “soft” vowels. They are opposed to the “tense” vowels a, �, � and �� and correspond to the “hard” vowels. There is only one phoneme � which is phonetically lax (“soft”) but operates as a neutral vowel in this variety of Orochon (Zhang, 1989). These facts, in Boer’s opinion, can serve as proof that the vowel harmony must be defined in Evenki – and probably in all Tungusic—on the basis of a single phonetic parameter, namely the position of the tongue root. She notes that there are probably no neutral vowels i and u—probably with a few exceptions—either in literary Evenki or in its dialects. In Xunke Orochon, as described by Zhang, only one neutral vowel i exists. In the two Khamnigan Evenki dialects Borzya and Urulyungui, which were described by J. Janhunen, the “soft”� and “hard” � have really merged. Originally “hard” word stems which contain the neutral but phonetically “soft” i, are now followed by suffixes with “soft” vowels (Janhunen, 1991). As for the Evenki vowel harmony, as Boer concluded, there still remain several open problems that need to be further investigated (Boer, 1996:131). Several other interpretations of the Tungusic vowel harmony can be added to those listed above. These analyzed by Gisaburo N. Kiyose who attempted to show that the phenomenon is highly complex (Kiyose, 1996:147-50). According to N. Poppe, the Tungus language has strict rules of vowel harmony which is not of palatal-velar nature. Several back vowels are followed by certain back or even front vowels. But they are never followed by certain other back vowels. The vowels i and u are neutral (Popper, 1965:185). V.Iv. Tzintzius believed that it was the opposition of the vowel a and o to the vowel � on which the Evenki vowel harmony was based. The vowel e � is followed by the vowels a or a,� because the sound changes e � < *ai and e � < *ia have occurred. The vowels i and u are neutral but each of them was derived diachronically from front and back vowels. This is the reason why
gorel-1.p65
93
4/17/02, 8:56 AM
94
part three
both a and � may follow the vowels i and u occurring in a word stem (Tzintzius, 1949:122). In accordance with V.A. Avrorin’s opinion, the opposition of the low vowels a, o and ι versus the high vowels �, u and i is the basis of vowel harmony in Nanai (Gold), but the low ι and the high i are rendered orthographically by the same letter (Avrorin, 1959:40). Analyzing all these interpretations of vowel harmony in the Tungusic languages, Kiyose suggested that it is the Korean data that can help to solve problems concerning Tungusic synharmonism. In early Middle Korean an opposition of back versus front vowels (the vowels a, o, u were opposed to the vowels e, ö, ü, with the vowel i being neutral) existed. As a result of several very important sound changes (e > � , o > a/�, ö > �, u > o, ü > u) this opposition collapsed in the course of time. In Kiyose’s view, the vowel harmony in Middle Korean was of palatal-velar nature (Kiyose, 1997:149). Vowel harmony is one of the basic phonetic phenomena of the Manchu sound system. However, Manchu vowel harmony differs from that of other Tungusic languages. By contrast, classical Manchu has two groups of suffixes. When attached to a word stem, those of the first group, change their sound value under the influence of stem vowels. Participial, derivational, and formbuilding verbal suffixes, as well as certain ones for plurality, belong to this group. Those of the second group, when appended to a word stem, do not change their vowels to harmonize with the stem vocalic sounds. They are markers for the accusative (be), dative-locative (de), and separative (deri) cases, as well as suffixes for the imperfect converb (-me) and nominalizers (-ngge, -ningge and ba). Gisaburo N. Kiyose explains this fact by the collapse of palatalvelar harmony in several suffixes in Manchu of the Ming period (1368-1644). In Jurchen of the Jin period (1115-1234) the usage of vowels in all suffixes was strictly ruled by vowel harmony which was of palatal-velar nature, that is the opposition of front versus back vowels, the front vowel e versus the back vowel a included. Then, with the shift of the sound value of the vowel e from the mid-front [e] to the mid-central, that is schwa [�], the back/front opposition was destroyed. As a result, several suffixes and markers listed above had lost their back counterparts. Here are some examples: jur. -mai/ -mei > ma. -me (the suffix for the imperfect converb); jur. -ba/-be > ma. -be (the marker for the accusative case). But several other suffixes had not only preserved the e/a opposition but added one more variant
gorel-1.p65
94
4/17/02, 8:56 AM
95
phonology
containing the rounded vowel o. It can be viewed as the appearance of the labial harmony which did not exist in Jurchen before, for example: -ha/-he/-ho, -hai/-hei/-hoi, -la/-le/-lo, -ra/-re/-ro (Kiyose, 1997:149-50). According to Möllendorff, the vowel harmony in Manchu can be considered as a certain attraction of vowels physiologically related to each other, when a vowel can be followed by the corresponding one. The majority of suffixes with regard to a vowel, which they contain, have a choice between two or three of vowels. If a stem attaches two or more suffixes, the vowel of the first suffix determines the vowels of the others (Möllendorff, 1892:2). Nevertheless, all the theories on vowels harmony in classical Manchu do not fully answer all questions about its nature. What we know is that in classical Manchu, the vowel harmony is realized as the opposition of the vowels a and o to the vowel � (written as e). All Manchu vowels can be distributed into three classes. The first class contains the vowels a, o and � (the so-called “hard” å which is also written as � � or o �). The second class consists of the single vowel � (written as e). The third class includes the vowels i and u (the socalled “soft” u). There is a strict requirement that vowels of the first and the second classes do not occur together within a word. A stem normally contains the same vowel as its first syllable, and it may be augmented only with suffixes including the same vowel. Alternatively, it may be augmented with a suffix containing a vowel of the same class to which the initial vowel of a stem is referred. Here are several examples of this type of so-called total synharmonism: aldangga “distant” (in relationship), eldengge “shining,” “glorious,” “glowing,” horonggo “powerful,” “majestic,” “regal,” ala-ha, ala-ra (ala- “to tell”), batala-ha, batala-ra (batala- “to be an enemy,” “to oppose”), ejele-he, ejele-re (ejele- “to be master of,” “to rule,” “to occupy by force”), genehe, gene-re (gene- “to go”), okdono-ho, okdono-ro (okdono- “to go to greet”), borhono-ho, borhono-ro (borhono- “to form a heap or swarm”), where the suffix -ngga/-ngge/-nggo, as well as the suffixes for the imperfect and perfect participles (-ra/-re/-ro and -ha/-he/-ho respectively) harmonize with a vowel of the initial syllable of a word. It is admissible that initial vowels of both classes, the first and the second, are followed by a vowel of the third class, that is vowels i and u (the so-called “soft” u). This type of vowel harmony has the name “step harmony.” The following examples demonstrate this type of synharmonism:
gorel-1.p65
95
4/17/02, 8:56 AM
96
part three
acabun “summary,” “union,” “adaptation,” “harmony,” dabali “excessively,” “too,” ejetun “record,” “description,” ergi “direction,” “side,” enduri “spirit,” “god,” “deity,” bontu “an adze,” orin “twenty.” It is also admissible that initial vowels i and u may be followed by vowels of the first or the second classes within a stem: ilan “three,” ilha “flower,” “blossom,” but igehe “the stem of fruits and melons;” isina- “to reach,” “to arrive,” sinda- “to place,” “to put,” “to set,” but kice- “to strive,” “to exert oneself,” “to be intent on,” “to study,” file- “to warm oneself by a fire;” ulana “a kind of a small, red, sour cherry,” funggala “tail feather,” “feather in an official’s hat,” munggan “low hill,” “mound,” “tomb,” but unde “not yet (particle used after imperfect participle),” undehen “rod,” “staff,” board,” fulehe “root,” muke “water,” “river,” “stream.” If a stem contains the vowels i or u, then appended suffixes may occur with vowels of both classes (the alternation a/e): aci-ha (aci- “to load”), ali-ha (ali- “to receive,” “to accept”), isi-ha (isi- “to pull up (grass),” muda-ha (muda- “to return”), tuwa-ha (tuwa- “to look,” “to look at,” “to see”), daru-ha (daru- “to buy on credit,” “to last,” “to endure”), but bi-he (bi- “to be,” “to exist”), ji-he (ji- “to come”), bu-he (bu- “to give), ulhi-he (ulhi- “to understand,” “to comprehend”), where the suffix for the perfect participle changes its vowel (-ha/-he). This fact may probably serve as an indication that the word stems were still belonged to the original pattern of vowel harmony by the time when they were recorded. Also there were two distinctive phonemes of each of the vowels i and u (“soft” and “hard” ones). If a stem begins with a vowel of the first class and ends in the vowel i, some appended suffixes may contain a vowel of the same class, while some may include the vowel � : taci-ha but taci-re (taci- “to learn,” “to study”), mari-ha but mari-re (mari- “return,” “to go back”), miyali ha but miyali-re (miyali- “to measure”), donji-ha but donji-re (donji- “to listen,” “to hear”), omi-ha but omi-re (omi- “to drink”), where the suffix for the perfect participle (-ha/-he/-ho) harmonizes with a vowel of the initial syllable of a word but the suffix for the imperfect participle (-ra/-re/-ro) does not, although it normally does. The so-called “soft” vowel u is mostly followed by the vowel i: usin “field,” usisi “farmer,” turi “bean,” “pea,” but there are some word stems in which the “hard” vowel å, its sound value is [�], is followed by the vowel i: gånin “thought,” “sense,” “mind;” gåli- “to be in good terms;” håri “pine nut.” There are some words which contain the “hard” vowel å and the “soft” vowel u: buhå “deer,” butån “crock,” “large jar,” jugån
gorel-1.p65
96
4/17/02, 8:56 AM
97
phonology
“road,” and gåsu “thick, heavy rope,” gåtucun “shame,” “insult,” “shameful,” håsun “strength’, “power,” håturi “good luck,” “good fortune.” There is one more fact that can serve as a confirmation of the existence of the opposition of the vowels a and o versus the vowel � in Manchu. The way the letters t and d are written, when they are followed by the letters a, o or i (rendering the vowels a, o or i respectively), differs from the way they are written when followed by the letters e, u (rendering the vowel � and the “soft” vowel u), and å (rendering the “hard” å which sound value is [�]). Bearing in mind that the original Manchu vocalic inventory presumably included eight vowels, four back vowels being opposed to four front ones, as viewed by a number of manjurologists, one could think that the original vowel harmony pattern in Manchu was based on the opposition of front versus back vowels, that is palatalvelar harmony. Synharmonism in Manchu was possibly similar to the vowel harmony in Jurchen, as it was understood by Kiyose. Subsequently this pattern shifted to a harmony based on another pattern, probably, on relative height, with the “soft” [u] rising higher than the “hard” [�], and the mid-central [�] rising higher than the low back [a]. However, many questions still remain open. The nature of vowel harmony in Manchu is not quite clear and it needs to be investigated further. It should be emphasized that in classical Manchu there was an extremely strict correlation between vowel harmony and the alternation of certain consonants. But this alternation itself was treated in different ways by specialists. Zakharov was of the view that there existed only three back consonants k, g, and h, each represented by two different graphemes, depending on the vowels which occurred with them. According to him, it is exactly the opposition of the vowels a, o to the vowel � (written as � in the Cyrillic alphabet) that determines the way of writing of the back consonants. The consonants k, g, h are rendered by different letters depending on the vowel which occurs with them within a syllable. The sound value of these consonants—which could be realized as either velar or uvular—is determined by the vowel with which they constitute a syllable. Using the terminology of contemporary phonetics, we can say that Zakharov considered the two series of these consonant sounds to be allophones occurring in complementary distribution with each other. Phonologically they
gorel-1.p65
97
4/17/02, 8:56 AM
98
part three
were accepted by him as the same phonemes. Zakharov noted that this fact was called by the Manchus themselves “kaharakca keherekce.” These two words have no particular meaning but display different combinations of consonants and vowels that could occur in words. According to him, there was a single vowel u which could occur both after the “hard” and “soft” variants of back consonants. Occurring after the “hard” consonants k, g, h (the uvulars [q], [�] and [�] in contemporary linguistics), the vowel å, written as y� in the Cyrillic alphabet, had a different sound value (the “hard” å, its sound value is considered to be [�] in contemporary linguistics) and different writing. In his view, the problem of the contrasting phoneme å appeared solely due to the fact that in Manchu the letters o and u are distinguished by a dot which was also the means of distinguishing the letter for the consonant g from the letters for the consonants k and h (Zakharov, 1879:51-2, 55-8). According to other specialists, there was an opposition between velar and uvular phonemes in classical Manchu, and the existence of two graphic variants for the vowel u, not two different phonemes, should be treated orthographically, as a special device for distinguishing preceding velars and uvulars (Ligeti, 1952). There is also an opinion according to which there are contrasting consonants, velars and uvulars, and contrasting phonemes the “soft” u with the sound value is [u], and the “hard” å with the sound value is [�] and “soft” [o], which occurs to render Mongolian borrowings (spelled as ö in Mongolian). In conclusion, it is a fact that in Manchu, there is a strict correlation between vowels of a certain class and a series of back consonants, velars or uvulars (no matter how they are viewed, as phonemes or allophones). As it was mentioned above, the harmony of vowels in Manchu is mainly restricted to the opposition of the vowels a and o to the vowel �. The velars are followed by the vowels �, while the uvulars are followed by the vowels a or o. There is also a strict correlation between the different letters for the “soft” vowel u and the “hard” vowel å and the two series of the letters for the back consonants. The letters for the velars are followed by the letter for the “soft” vowel u, while the letters for the uvulars are followed by the letter for the “hard” vowel å. It is an open question, whether these letters represent different phonemes (or allophones) or they simply serve to distinguish velars from uvulars which, in turn, can be treated as allophones occurring in complementary distribution.
gorel-1.p65
98
4/17/02, 8:56 AM
99
phonology
It is also partly a question of basic assumptions on which the phonetic analysis is based. The way of writing the alveolar consonants t and d also depends on which one of the vowels, a, o, or �, occurs with them.
3. The Dropping of Vowels and Consonants in Manchu Words Vowels are often omitted in the middle of words. Here are some instances: butha “hunt” < butaha; cirku “pillow” < ciruku; forgon “season,” “fate” < forogon; gelhun “fear,” “timid,” “fainthearted” < gelehun; hojhon “son-in law” < hojihon (hojigon); ilha “flower” < ilaha; ufhi “part,” “share,” “portion” < ufuhi. When two words combine, in the resulting ones vowels or the second element of diphthongs are often dropped: aba “where?” < ai “what?,” “which?” + ba “place;” amargi “back,” “behind,” “north” < ama(la) “behind” + ergi “side,” “direction;” erse = PL of ere “this” < ere + se which is the suffix for plurality; ertele “up till now” < ere “this” + tele which is the suffix for the converb denoting the meaning “up to now.” Consonants are usually omitted in the intervocalic position in a word: cu(w)an “boat,” “ship;” gå(w)a “other,” “another;” ju(w)an “ten;” kå(w)a “light-yellow (horse).” The consonants k and h, g and h may be interchanged, with dropping vowels: emhe “mother-in-law (wife’s mother, sometimes a husband’s mother)”—emeke “mother-inlaw (husband’s mother);” julge “ancient times,” “antiquity”—julehe “formerly” (Möllendorff, 1892:3).
4. Stress The stress of Manchu words has not been investigated in a satisfactory manner. According to Zakharov who actually heard the spoken language, the stress was not governed by strict rules in Manchu. Basic stress rules according to Zakharov are outlined below (Zakharov, 1879:62-4). If a word consists of a few open syllables (the last one may be closed), then all of them are pronounced with equal strength, accompanied by a lengthening of the initial vowel. If a word consists of a few syllables, when the first is closed or contains any of the descending diphthongs ai, ei, oi, ui, eo and the second begins with any consonant, then the entire word seems to be divided into two
gorel-1.p65
99
4/17/02, 8:56 AM
100
part three
parts, each of them pronounced as a new word with equal strength and a lengthening of the initial vowel. In noun morphology the suffixes affect the stress of the words to which they are attached in the following ways: 1) The stress always falls on the case marker which follows the word: alin i � dorgide “inside a mountain,” booci � tucike “(somebody) went from a house.” 2) When nouns are appended with any of the interrogative particles, reduced to the n, the stress falls on the final syllable: saiyu� n (interrogative form of sain): si saiyå� n? “How are you?;” yargiyå� n (interrogative form of yargiyan) “Is it true?” 3) If an adverb is derived from a noun by means of the form of the genetive case, then the stress falls on the case marker: cun cun i� “gradually,” “by degrees,” saikan i � “rather well.” 4) The stress falls on the suffixes -kan /-kon/-ken attached to a noun to express the diminutive meaning: elhe “calm,” “gentle,” “quiet,” “slow,” “well-being”—elheke�n “rather well, gentle,” “rather slow,” golmin “long” —golmika�n “rather long,” sain “good,” “well”—saikan “rather well,” “nicely.” In verb morphology the suffixes and particles also affect the stress of the words to which they are attached: 1) The perfect participle form (in -ha/-ho/-he) is characterized by the basic stress rule, that is even strength applied to all syllables of a word with a lengthening of the initial vowel. But in the converbal form in -hai/-hoi/-hei which is obviously derived from this participle and has the frequentative and durative meanings, the stress falls on the last syllable: alahai� “(somebody) told repeatedly, many times;” tehei� “(somebody) sat for a long time.” 2) The imperfect participle form (in -ra/-ro/-re) has usual stress, but when it is used to express the future, the stress falls on the last syllable: urunakå alara � “(I) shall tell certainly, surely.” 3) The prohibitive form which is derived from the imperfect participle by means of the particle ume placed before the participial form and used for negation, also has the stress on the last syllable:
gorel-1.p65
100
4/17/02, 8:56 AM
101
phonology ume alara � “do not tell.”
4) In a special Manchu form conveying the meaning of apprehension (-rahå), the stress also falls on the last syllable: alarahå� “there is an apprehension that (one) will tell (it).” 5) In the imperative form which coincides with a verbal stem, the stress always falls on the last syllable: ala � “tell!;” alabu� “order (somebody) to tell;” alana� “go and tell.” 6) When the synthetic desirative form in -ki is used to denote the future, the stress falls on the last syllable which is -ki: alaki � “(I) shall tell.” 7) In the analytical form which is derived from the synthetic desirative form in -ki and the stem of the verb se- “to say,” the stress always falls on the last syllable of the first word of this analytical construction: alaki� sembi “I wish to tell.” 8) The stress usually falls on the interrogative particles ni, nio attached to verbal forms. When these particles are reduced to n occurring in the negative forms of the imperfect and perfect participles (-rakån, -hakå�n) or replaced by the particle o, pronounced as [u], the stress falls on the last syllable of the verbal form: ainu balai gisurembini � ? “Why do (you) say (it) unreasonably?;” yargiyan i � semeo� ? “Is it true what (you) are talking?;” alahao � ? “Did (you) say (that)?” 9) Onomatopoeic words always have the stress on the last syllable: koto�r kata� r (onom.) = the sound of hard things rattling together or falling, the sound of a flock of pheasants flying; halar� hilir� (onom.) = the sound of bells on the girdle, the sound made by a shaman’s sword; kalàng kiling � (onom.) = the sound of metal or stone objects banging together; pekte� pakta � “perplexed,” “dismayed.” 10) When a lexical or grammatical meaning is expressed by a pair of close synonyms, both words have the stress on their last syllable: etuku � adu � “clothing,” agura � tetu� n “tools,” oilori� deleri� “superficial,” “trivial,” “frivolous;” acu� facu� “with loving tenderness.” Möllendorff noted that the stress falls always on the last syllable of a word, and regarded it as a similarity with Mongolian (Möllendorff, 1892:1). Norman presented more data concerning the dialect he described.
gorel-1.p65
101
4/17/02, 8:56 AM
102
part three
According to him, the stress in Sibe is by and large predictable and it falls usually on the penultimate vowel or diphthong within the word. In a few cases where the penultimate vowel is /i/, /u/ or /�/, the stress falls on the antepenultimate vowel. The suffixes affect the stress of the words to which they are attached in two ways. If the suffix is monosyllabic and begins with a voiced sound or with one of the lenis stops, then the stress falls on the penultimate vowel or diphthong of the word. If the suffix is disyllabic or monosyllabic and begins with a voiceless sound or one of the tense stops, the stress of the resulting word is similar to that in the unsuffixed stem. In some cases disyllabic suffixes may have secondary stress on the first syllable (Norman, 1974:164).
5. Types of Syllables and Syllabic Structure of Manchu Words In accordance with the syllabic character of Manchu words, no one word in Manchu commences with two (or more) consonants or ends in them (I use the term syllable to refer to a phonemic unit, not a phonetic one). Two consonants may occur in the middle of a word, because a preceding syllable may end in a consonant and a following one begin with a consonant: ak-dan (akdan “trust”), dur-dun (durdun “crepe”), dal-ba (dalba “side”), mek-ten (mekten “bet,” “wager”), or-hoda (orhoda “ginseng”). Not one native Manchu word commences with the vowel å, as in the Mongolian borrowings ålet and åren, or with the consonant r which occurs in Buddhist works in some transliterated Sanskrit words. Final m, l and t are only found in foreign words, for example: mandal “the mandala (Buddhist) < mo. mandal “a special place where sacred rites are practiced” < skr. mandala. Final k, r and s occur only in onomatopoeic words: kåwak cak (onom.) = the sound of fighting with poles or sticks; fak fik (onom.) = the sound of fruit falling; kalar kilir (onom.) = the sound of keys or small bells jingling; sir siyar (onom.) = the sound of grass and leaves moving slightly; kåwas (onom.) = the sound of chopping wood, the sound of a falcon striking an object with its wings; pes pas (onom.) = the sound of something soft ripping. Final � also occurs in onomatopoeic words, like in gang ging (onom.) = the sound of a flock of wild geese calling; tang tang (onom.) = the sound of a bell; jing yang (onom.) = the sound of birds singing
gorel-1.p65
102
4/17/02, 8:56 AM
103
phonology
harmoniously. The final consonants k, r and s occur in words borrowed from Chinese as well. A few words end in the consonant b: tob “right,” “straight,” “upright;” cob seme “appearing suddenly,” “standing out from the crowd;” kab kib seme “snapping at each other (of dogs fighting or biting);” tab tib (onom.) = the sound of dripping water. There are two types of syllables in Manchu, open and closed. The first pattern of an open syllable may consist of a vowel, a geminate vowel or a diphthong (V), and the second pattern includes a consonant followed by a vowel or a geminate vowel or a diphthong (CV). A closed syllable may be presented by a vowel with a following consonant (VC), or may begin with a consonant followed by a vowel and a consonant (CVC), including the consonant n (CVn). According to Hattori & Yamamoto, in the dialect they described there is a voiceless glottal catch (laryngeal) which precedes all vowels in initial position. In this way of thinking, there are no syllables beginning with a vowel component in Manchu.This point of view is not confirmed by other specialists. However, it is known that the Proto-Tungusic *x- disappeared in Manchu: *x- > ma. ø- (Benzing, 1955:990-1). According to Kiyose, the initial syllable ha- [�a] in Jin Jurchen appeared as the vowel a- in Ming Jurchen and further in classical Manchu, losing its consonant: *hadu “dress” in Jin Jur. > *adu id. in Ming Jur. > ma. adu id; *habka “heaven” in Jin Jur. > *abka id. in Ming Jur. > ma. abka id. This laryngeal, postulated by Hattori & Yamamoto, could be a reflex of the Proto-Tungusic *x(Kiyose, 1995). It is common for verbal forms to contain a lot more syllables that nominal words do. This fact can be explained by the grammatical nature of the verbal forms which may display several morphological categories synthetically, so that the verbal stem may be augmented with a number of morphemes. It should be taken into account, however, that the syllabic structure of a word is not identical to the morpheme division, therefore the syllabication is irreconcilable with morphemic analysis. In Literary Manchu the syllabic structure of words displays the following patterns. 1) Monosyllabic words: V a “the male or positive principle,” “yang” e “the female or negative principle,” “yin”
gorel-1.p65
103
4/17/02, 8:56 AM
104
part three ai “what?” sib. /�aa/ [�a�]
da “root,” “base;” sib. /daa/ [da' ]
bi “I,” “me” (personal pronoun for 1.SG); sib.
/bii/ [bi']
moo “tree,” “wood,” “timber,” “lumber,”
“log”
bai “plain,” “simple,” “only”
cai “tea;” sib. /ciaa/ [t���]
deo “younger brother, “younger (of male)”
den “tall”
duin “four”
fiyan “rouge,” “luster,” “appearance;” sib.
/fiaN/ [f���n]
CV
CVn
2) Disyllabic words: V-CV
V-CVn VC-CV VC-CVn
CV-CV
CV-CVn
gorel-1.p65
104
aba “hunt,” “battue;” sib. /�abaa/ [�aba�]
ehe “evil,” “wicked;” sib. /�ex�/ [��x]
aika “if;” sib. /���qaa/ [�aιqa�]
oori “semen,” “spirit,” “essence,” “energy”
aniya “year;” sib. /�ani/ [���]
efin “game,” “sport;” sib. /�ifiN/ [�ifin]
aisin “gold,” “golden;” sib. /���s�iN/ [��ι�ι��
abka “sky,” “heaven;” “weather;” sib.
/�f�qaa/ [�afqa�]
arbun “appearance,” “form;” sib. /����v�N,
�a��vuN/ [�arv�n, �arvun]
elgiyen “good harvest,” “good year;”sib.
/�elixiN/ [��lj�in]
buda “food,” “cooked cereal,” “cooked rice;”
sib. /b�daa/ [b�da�]
gebu “name;” sib. /gev�/ [��v]
cooha “army,” “troops,” “soldier,” “military;”
sib. /cuah�/ [t�ω� ��]
leose “tower,” “building;” sib. /le�us�/ [l�us]
meihe “snake,” “serpent;” sib. /me�ix �/
[m�ιx]
kemun “measure,” “model,” “rule,” “ruler”
caliyan “pay and provisions (military)”
4/17/02, 8:56 AM
105
phonology
meifen “neck” doohan “bridge” katak (onom.) = the sound of a lock clicking shut funde “instead of” tondo “straight,” “frank;” sib. /toNd�/ [t�nd] bandan “backless bench” bithe “book,” “document;” sib. /bit� xee/ [bitx��] solho “Korea,” “Korean;” sib. /sol�hoo/ [s�����] sargan “wife;” sib. /sar�h�N/ [sar��n]
CV-CVC CVn-CV CVn-CVn CVC-CV
CVC-CVn 3) Trisyllabic words: V-CV-CV
V-CVn-CV V-CVC-CV V-CVC-CVn
V-CV-CVn
Vn-CV-CV Vn-CV-CVn
Vn-CVn-CV
gorel-1.p65
105
agåra “tool,” “implement,” “weapon;” sib. /�ahur�/ [���ωr] aihåma “tortoise (female),” “turtle (female);” sib. /���hum�/ [�ai�ωm] oolame = CONV in -me from oola- “to step aside,” “to make way”) ubiyada “disagreeable;” sib. /�uviad�/ [�uv���d] alinji- “come to accept;” sib. ali- /�aialim�/ [��lim] “to accept” amargi “back,” “behind,” “north;” sib. /�am�rixi/ [�am�rj�j] ajirgan “a male horse, donkey, camel, or dog;” sib. /�aj�r �haN, �ajir�haN/ [������� an, ����ιr�an] “stallion” aibishån “slightly swollen” amuran “fond of,” “intent on,” “afterbirth;” sib. /�amuraN/ [�amuran] aliyacun “regret,” “waiting” enduri “spirit,” “god,” “deity;” sib. /�eNduri/ [��ndurj] oncohon “facing upward;” sib. / �oNc�h�N, �oNcuhuN/ [���������, �������ωn] anculan (giyahån) “hawk” uncanji = IMP from uncanji- “to come to sell;”
4/17/02, 8:56 AM
106
part three
Vn-CVn-CVn Vn-CVC-CV Vn-CVC-CVn VC-CV-CV VC-CV-CVn VC-CVn-CV VC-CVn-CVn VC-CVC-CV CV-CV-CV
CV-CVn-CV CV-CV-CVn
CV-CVn-CVn CV-CVC-CV CV-CVC-CVn CVn-CV-CV CVn-CV-CVn CVn-CVn-CV CVn-CVC-CV CVn-CVC-CVn CVC-CV-CV CVC-CV-CVn CVC-CVn-CV
gorel-1.p65
106
sib. unca- /�uNcam�/ [�u���am] “to sell” ------andargi “neighbour,” “neighbouring,” “adja cent” undustan “Hindustan” ambula “greatly,” “widely,” “very much” arsalan “lion;” sib. /�ar�s�l�N/ [�ars�l�n] amhålan “a whistle” urgunje- “to be glad,” “to rejoice;” sib. /�uruxuNj�m�/ [�ur�������m] ------amgambi = IMPF in -mbi of amga- “to sleep” muduri “dragon” boigoji “master;” sib. /bo�ihoji/ [b�ι����] funiyehe “hair (of the head)” suwanda “garlic” doigonde “in advance,” “beforehand;” sib. /do��h�Nd�/ [d�ι���nd] kåwaran “courtyard;” sib. /quar��� [qω� ar�n] beikuwen “frost,” “cold” muhaliyan “ball,” “sphere,” “pile,” “stack” ----banambi = IMPF in -mbi of bana- “to get,” “to obtain,” “to be able” somishån “hidden,” “secret;” sib. /s�iomis�huN, siomis�huN/ [��mis�ωn, s�mis�ωn] fontoho “tunnel,” “hole” kandahan “Manchurian elk” banjinji = IMP of banjinji- “to come to live (in a new place)” donjimbi = IMPF in -mbi of donji- “to listen,” “to hear” banjishån “having sufficient money or goods to lead a comfortable life” debsiku “a fan (made of feathers);” sib. /def�s�k�/ [d�fsk] debtelin “volume;” sib. /def�t�liN/ [d�ft�liN] korsondu- “to be mutually annoyed,” “to regret mutually”
4/17/02, 8:56 AM
107
phonology CVC-CVn-CVn CVC-CVC-CV
CVC-CVC-CVn
-----tuksimbi = IMPR in -mbi of tuksi- “to pound,” “to throb (of the heart),” “to be afraid,” “to be anxious” dekderhån “one name for the seagull”
4) Words of four syllables: V-CV-CV-CV asihata “young man;” sib. /��� ihat�/ [� ���at] V-CV-CV-CVn uhukeliyen “rather weak, soft” V-CVn-CV-CV alinjime = CONV in -me of alinji- “to come to accept;” sib. ali- /�ialim�/ [��lim] “to accept” V-CVn-CVC-CV alinjimbi = PERF in -mbi of alinji- “to come to accept;” sib. ali- /�ialim�/ [��lim] “to accept” V-CVC-CV-CV alimbaha- “to be able to bear” V-CV-CVC-CV abalambi = PERF in -mbi of abala- “to participate in a battue,” “to hunt;” sib. abala- /�av�l�m�/ [�av�l�m] VC-CV-CV-CV abkawaru “cursed by heaven,” “an oath” VC-CV-CVC-CV abtalambi = PERF in -mbi of abtala- “to break off (branches),” “to prune” VC-CV-CV-CVn ambulakan “rather greatly,” “rather much” VC-CVC-CV-CV aààandume = CONV in -me of aààandu- “to move together” VC-CVC-CVC-CV aààandumbi = IMPF in -mbi of aààandu- “to move together” CV-CV-CV-CV simacuka “lonely,” “desolate,” “scant” wacihiyame “completely,” “wholy;” sib. /vacixiam�/ [vat�x��am] hairacuka “pitiable,” “pitiful,” “too bad;” sib. /ha�iracuqu/ [��ιrat�uqw] CV-CV-CV-CVn fudasihån “rebellious,” “disloyal,” “obsti nate” CV-CVn-CV-CV baicanjime = CONV in -me of baicanji- “to come to inspect” CV-CVn-CVC-CV baicanjimbi = IMPF in -mbi of baicanji- “to come to inspect”
gorel-1.p65
107
4/17/02, 8:56 AM
108
part three
CV-CVC-CV-CV
CV-CVC-CVC-CV
CV-CV-CVC-CV CV-CVC-CVC-CV
CVn-CV-CV-CV CVn-CV-CVC-CV CVn-CVn-CV-CV CVn-CVn-CVC-CV CVC-CV-CV-CV
CVC-CV-CVC-CV CVC-CVn-CV-CV CVC-CVn-CVC-CV CVC-CVC-CV-CV CVC-CVC-CV-CVn CVC-CVC-CVC-CV
5) Words of five syllables: V-CV-CV-CV-CV
V-CV-CV-CVC-CV
gorel-1.p65
108
tebumbihe = PAST in -mbihe of tebu 1. CAUS of te- “to sit,” “to live,” “to
occupy;” 2. “to pour,” “to plant,” “to
put in,” ect.
hiyahaljambi = IMPF in -mbi of hiyahalja-
“to cross one another,” “to cross back
and forth”
gelebumbi = IMPF in -mbi of gelebu- =
PASS of gele-“to fear,” “to be afraid”
barambumbi = IMPF in -mbi of barambu
1. CAUS in -mbu of bara- “to mix
together,” “to pour soup on rice;”
2. “to mix,” “to mix among,” “to
mingle together”
fancacuka “annoying,” “causing despair”
banjibungga “productive,” “creative”
banjinjime = CONV in -me of banjinji “to
come to live (in a new place)”
banjinjimbi = IMPF in -mbi of banjinji
korsocuka “regrettable,” “annoying;”
sib. /qor�secuq�, qor�sucuqu/
[q�rs�t�uq‘, q�rsut�uqw]
bektelembi = IMPF in -mbi of bektele- “to
cut off the feet as a punishment”
baktandame = CONV in -me of baktanda-
“to contain,” “to hold”
baktandambi = IMPF in -mbi of baktanda-
“to contain,” “to hold”
bardanggila- “to boast”
duksursehun “rough (of terrain)”
dabduràambi = IMPF in -mbi of dabduràa-
“to flare up,” “to go into a rage”
ahåcilame = CONV in -me of ahåcila- “to treat as one’s senior,” “to be older than” ahåcilambi = IMRF in -mbi of ahåcila“to treat as one’s senior,” “to be older than”
4/17/02, 8:56 AM
109
phonology
V-CVC-CV-CV-CV
V-CVC-CV-CVC-CV VC-CV-CV-CV-CV VC-CV-CV-CVC-CV VC-CV-CVn-CV-CV VC-CV-CVn-CVC-CV CV-CV-CV-CV-CV
CV-CV-CVn-CV-CV CV-CV-CVn-CVC-CV CV-CV-CV-CVC-CV CV-CV-CVn-CV-CV CV-CV-CVn-CVC-CV CV-CV-CVC-CV-CV
CV-CV-CVC-CVC-CV CV-CVC-CV-CV-CV
gorel-1.p65
109
aisilabukå “helper,” “prime minister” ainambahame = CONV in -me of ainambaha- “how to obtain?,” “how can?” ainambahambi = IMPF in of ainambaha“how to obtain?,” “how can?” albatulame = CONV in -me of albatula“to act or speak coarsely” albatulambi = IMPF in -mbi of albatula“to act or speak coarsely” alhådanjime = CONV in -me of alhådanji- “to come to imitate” alhådanjimbi = IMPF in -mbi of alhådanji- “to come to imitate” bederebume = CONV in -me of bederebu 1. PASS/CAUS of bedere- “to re turn,” “to withdraw (at court or at a ceremony),” “to die (of a noble per sonage);” 2. “to send back,” “to withdraw,” “to refuse,” “to return a courtesy or gift” banihånjame = CONV in -me of banihånja- “to treat kindly” banihånjambi = IMPF in -mbi of banihånja- “to treat kindly” bederebumbi = IMPF in -mbi of bederebu hebeàenjime = CONV in -me of hebeàenji“to come to discuss” hebeàenjimbi = IMPF in -mbi of hebeàenji- “to come to discuss” badarambume = CONV in -me of badarambu- 1. PASS/CAUS of badara“to become wide,” “to expand,” “to become larger,” “to become pros perous;” 2. “to enlarge,” “to ex pand,” “to propagate” badarambumbi = IMPF in -mbi of badarambu fataràabume = CONV in -me of
4/17/02, 8:56 AM
110
part three
CV-CVC-CV-CVC-CV CV-CVC-CVC-CV-CV
CV-CVC-CVC-CVC-CV CVn-CV-CV-CV-CV
CVn-CV-CV-CVC-CV CVC-CV-CV-CV-CV
CVC-CV-CV-CVC-CV CVC-CV-CVn-CV-CV CVC-CV-CVn-CVC-CV CVC-CVC-CV-CV-CV CVC-CVC-CV-CVC-CV
6) Words of six syllables: V-CV-CV-CV-CV-CV
V-CV-CV-CV-CVC-CV V-CVC-CV-CV-CV-CV V-CVC-CV-CV-CVC-CV
gorel-1.p65
110
fataràabu- CAUS. of fataràa- “to pinch repeatedly,” “to be in the habit of using things sparingly” fataràabumbi = IMPF in -mbi of fataràabu hafiràambume = CONV in -me of hafiràambu- PASS/CAUS of hafiràa“to use economically” hafiràambumbi = IMPF in -mbi of hafiràambu fondolobume = CONV in -me of fondolobu- PASS/CAUS of fondolo“to penetrate,” “to go through” fondolobumbi = IMPF in -mbi of fondolobu- PASS/CAUS of fondolo bakcilabume = CONV in -me of bakcilabu- 1. PASS/CAUS of bakcila- “to rear (of horses),” “to come to a sudden stop;” 2. “to put in opposition,” “to make oppose” bakcilabumbi = IMPF in -mbi of bakcilabu hargaàanjime = CONV in -me of hargaàanji- “to come to court” hargaàanjimbi = IMPF in -mbi of hargaàanji- “to come to court” bardanggilame = CONV in -me of bardanggila- “to brag,” “to boast” bardanggilambi = IMPF in -mbi of bardanggila
anafulabume = CONV in -me of anafulabu- PASS/CAUS of anafula“to garrison,” “to guard a frontier” anafulabumbi = IMPF in -mbi of anafulabualimbanarakå “intolerable,” “insup portable,” “greatly,” “exceedingly” ilimbahabumbi = IMPF in -mbi of
4/17/02, 8:56 AM
111
phonology
Vn-CV-CV-CV-CV-CV Vn-CV-CV-CV-CVC-CV CV-CV-CV-CV-CV-CV
CV-CV-CV-CV-CVC-CV CV-CV-CV-CVn-CV-CV CV-CV-CV-CVn-CVC-CV CV-CV-CVC-CV-CV-CV CV-CV-CVC-CV-CVC-CV CVn-CV-CV-CV-CV-CV
CVn-CV-CV-CV-CVC-CV CVn-CV-CV-CV-CV-CV
CVn-CV-CV-CV-CVC-CV CVC-CV-CV-CV-CV-CV CVC-CV-CV-CV-CVC-CV
gorel-1.p65
111
ilimbahabuPASS/CAUS of ilimbaha- “to become accustomed,” “to get used,” “to be at peace with,” “to be calm” anduhårilame = CONV in -me of anduhårila- “to treat a person coldly” anduhårilambi = IMPF in -mbi of anduhårilabulekuàebume = CONV in -me of bulekuàebu- PASS/CAUS of bulekuàe“to look in a mirror,” “to reflect,” “to mirror,” “to perceive clearly” bulekuàebumbi = IMPF in -me of bulekuàebugucihiyerendume = CONV in -me of gucihiyerendu- “to be jealous of one another” gucihiyerendumbi = IMPF in -mbi of gucihiyerendu- “to be jealous of one another” deribunggileme = CONV in -me of deribunggile- “to make a beginning” deribunggilembi = IMPF in -mbi of deribunggile- “to make a beginning” gencehelebume = CONV in -me of gencehelebu- PASS/CAUS of gencehele“to strike with the back of a sword or like object,” “to land on the back” gencehelebumbi = IMPF in -mbi of gencehelebubontoholobume = CONV in -me of bontoholobuPASS/CAUS of bontoholo- “to be empty,” “to be bare,” “to be deprived” bontoholobumbi = IMPF in -mbi of bontoholobugåldarakålame = CONV in -me of gåldarakåla- “to get colic” gåldarakålambi = IMPF in -mbi of gåldarakåla- “to get colic”
4/17/02, 8:56 AM
112
part three
6. Elements of Morphophonemics. Morphemic Structure of a Word. Stems and Suffixes The morphological structure of the Tungus-Manchu languages is defined by most specialists as primarily suffix-agglutinative. The agglutinative ways of amalgamating morphemes within a word form clearly predominate over other means of word formation in Tungusic. In the Southern Tungusic languages agglutination is also of great importance. In comparison with other languages of the family, classical Manchu employs analytical devices in its morphological structure a lot more. Nevertheless, the agglutinative characterictics remain fundamental. A word consists minimally of a stem which may be simplex or derivative. In that case, a word coincides with a stem. The latter may be followed by one or more suffixes. A morpheme is defined here as a minimal language element possessing certain lexical or grammatical meaning. There is only one type of bound morpheme in Manchu. It is a suffix which may be attached directly to a nominal or verbal stem. Normally a suffix has monosyllabic or disyllabic structure, correspondingly, of CV or CVCV type. There are a few suffixes which commence with two consonants, but they are not numerous. They clearly originate from two morphemes amalgamated into a single one. For instance, the suffix for the imperfect finite form in -mbi was probably amalgamated from the suffix for the imperfect converb in -me and the copula bi “to be,” “to exist” (-mbi < -me + bi). Normally all stems remain stable when derivational or grammatical suffixes are added onto them. But there are some exceptions to this rule. The final -n in which some nouns end, is deleted before the plural suffix: amban “high official”—ambasa PL of amban; sadun “father of the son-in-law,” “father of the daughter-in-law”—sadusa PL of sadun; nikan “Chinese”—nikasa PL of nikan; irgen “people,” “the common people”—irgese PL of irgen; sargan “wife,” “woman,” “female”—sargata PL of sargan. In some stems, primarily derivative ones, which are augmented with suffixes, a vowel may be omitted. Thus, the word ertele “up till now,” “up to this point” is often used instead of eretele (ere “this”). The following verbal stems bederce- “to retreat,” “to withdraw,” ofordo“to incite by slander,” “to engage in malicious gossip;” takåràa- “to employ as a (personal ) servant” are mostly used instead of be
gorel-1.p65
112
4/17/02, 8:56 AM
113
phonology
derece-, oforodo-, takåraàa-. A stem may be augmented with a string of suffixes. Normally, not many suffixes may be added to a noun stem, because only a few nominal grammatical categories are expressed morphologically, with the help of suffixes. A verbal form may contain many more suffixes so that a verb stem may be appended with quite a long string of suffixes numbering from one to several of them (see Section 5). Normally there is no morpheme concatenation on the morphemic boundaries, with one exception, when the irregular pronominal case form occurs in the accusative. The final -n in which all irregular pronominal stems end, with the exception of the inclusive form for “we” that is muse, assimilates to a following consonant that is the consonant b in place of articulation: mimbe < min- “I” + be; simbe < sin- “you (SG)”+ be; imbe < in- “he,” ”she ” + be; membe < men- “we (EXCL)” + be; su(w)embe < su(w)en- “you (PL)” + be; cembe < cen“they” + be. There is one example of this rule applying to a noun in old Manchu: gisumbe < gisun “word,” “language” + be (Zakharov, 1879:133). In the Sibe dialect this rule applies to noun stems ending in -n. The final -n of a noun stem assimilates to a following consonant in place of articulation as it does before the accusative case in irregular pronominal case forms: morimbe < morin “horse” + be, alimbe < alin “mountain” + be, gaàambe < gaàan “village,” “country” + be; saihambe < saihan “beauty” + be (Norman, 1974:166; Lebedeva & Gorelova, 1994:48). 6.1. Noun Stems Noun stems may end in a vowel, a geminate vowel, a diphthong or final -n: abka “sky,” “heaven,” “weather;” ba “place,” “locality,” “spot;” da “root,” “base;” fa “window;” na “earth,” “ground;” ama “father,” “head of the household;” eme “mother;” gufu “husband of father’s sister;” uce “door;” use “seed,” “egg (of an insect);” àabi “disciple,” “student;” coko “chicken,” hiyehe “duck;” biya “moon,” “month;” beye “body,” “self;” oyo “top,” “roof;” uyu “turquoise;” uyuri “a black cat;” boo “house,” “family;” coo “a spade;” moo “tree,” “wood,” “stick,” “pole;” ai “what?,” “which?;” deo “younger brother;” geo “mare;” amban “high official,” “dignitary;” boljon “wave;” cihan “desire;” eten “force,” “resistence;” meifen “neck;” kumun “music;”
gorel-1.p65
113
4/17/02, 8:56 AM
114
part three
kåwaran “courtyard,” “camp (military),” “workshop,” “market place;” oyon “peak;” nioron “rainbow;” sihiyan “porch,” “pavilion.” In noun stems the final -n is deleted before the addition of certain suffixes, the plural, for instance: sargan “wife”—sargata “wives,” ahån “elder brother”—ahåta “elder brothers,” irgen “people,” “nation”—irgese “peoples,” “nations.” For most words the distinction “a simple—a geminate final vowel” is not noted in dictionaries of classical Manchu. By contrast, the dictionary of spoken Manchu (which is definitely the Sibe dialect) gives both phonemic and phonetic transcription where these distinctions are shown: abka /�af�qaa/ [?afqa'], ba /baa/ [ba'], da /daa/ [da'], fa /faa/ [fa'], na /naa/ [na'], ama /�amee/ [?am], te /tee/ [t�'], fe /fee/ [f�'], boro /boroo/ [boro'], coko /coqoo/ [t��q�'], biya /biaa/ [bia'], bi /bii/ [bi'], baju /bajuu/ [badzu'], gufu /gufuu/ [gufu'] (Yamamoto,1969). Norman also mentions some words ending in a geminate vowel in the Sibe dialect he investigated although in his dictionary of classical Manchu they are cited as those ending in a simple vowel. According to him, the second element of a geminate vowel is always deleted before a suffix (Norman, 1974:164-5). In some groups of nouns common morphemes can be discovered. These morphemes reveal their formal similarity with verbal ones. Thus, some nouns end in the morphemes -la/-le/-lo, -li: unggala “hole,” “cavity,” funggala “tail feather,” “feather in an official’s hat;” niyamala “moss found on trees and stones;” senggele “rooster’s comb;” sele “iron;” omolo “grandson;” konggolo “the crop of a bird;” holo “false;” kotoli “sail (of a ship);” hefeli “belly.” Nouns may end in the morphemes -ra/-re/-ro, -ri, -ru. The first three of them are formally similar to the suffix of the imperfect participle: duwara “a kind of fish (according to Zakharov “burbot;” according to Norman “mayfish,” “sweetfish”); bira “river;” sefere “a handful,” “a bundle;” nure “(rice) wine;” oforo “nose;” tohoro “circle,” “wheel;” toro “peach;” tomoro “a cup,” “a rather large bowl;” singgeri “rat,” “mouse;” mederi “sea;” juwari “summer;” bolori “autumn,” “fall;” tuweri “winter;” cungguru “navel.” Nouns may end in the morphemes -ka/-ke, -ha/-he and -ga/-ge which are formally similar to the suffix of the perfect participle: abka “sky,” “heaven;” ecike “father’s younger brother (uncle);” muke “water,” “river;” aciha “pack,” esihe “scale (of a fish);” erhe “green frog;” hasaha “scissors,” “shears;” ilha “flower,” “blossom,” nimaha “fish;” niyehe “duck;” usiha “star;” age “prince, son of an emperor,” “a polite
gorel-1.p65
114
4/17/02, 8:56 AM
115
phonology
term of address, master, sir, lord;” mejige “news,” “information;” ogo “mortar.” A group of nouns can be distinguished by the final morphemes -ma/-mo/-me, -mi, -mu: niyalma “man,” “person,” hasima “Manchurian crayfish,” ujima “domestic animal,” “livestock,” dehema “husband of mother’s sister (uncle),” deheme “mother’s younger sister (aunt),” erdemu “power,” “virtue,” oromo, oromu “cream,” namu “ocean,” “sea,” enggemu “saddle,” anami “a grown Manchurian elk”6. There is a small group of nominal stems which end in the following suffixes: -ba: dulimba “middle,” “centre;” hondoba “a kind of foxtaillike grass that can be eaten by horses;” -ji, -ju: boigoji “host,” “landlord,” “master;” boihoju “the god of the earth,” “the shrine of the earth god;” -hi, -ki: aduhi “leather trousers;” hålhi “muddleheaded,” “confused,” “blurred;” jabàaki “good fortune,” “luck,” “lucky;” lekerhi “sea otter;” -nggi: bardanggi “braggart;” etenggi “strong,” “powerful,” “hardy;” inenggi “day;” lebenggi “swampy,” “marshy,” “damp,” “muddy;” obonggi “bubble,” “foam;” nimanggi “snow;” nimenggi “oil,” “fat;” silenggi “dew.” 6.2. Verbal Stems Verbal stems may end in the following ways: 1) a single vowel: da- “to burn,” “to blow (of the wind),” “to rain,” “to snow,” “to take care of,” “to help,” “to work;” sib. /dam�/ [dam]; je- “to eat;” sib. /jem�/ [d��m]; ji- “to come;” sib. /jim�/ [d�ιm]; si- “to stop up,” “to fill in,” “to stand in for,” “to bribe;” sib. /s�im�/ [�ιm]; bu- “to give;” sib. /bum�/ [bum]; ku- “to swell,” “to bloat;” afa- “to attack,” “to make war;” baha- “to get,” “to obtain,” “to be able;” forgoào- “to turn,” “to rotate,” “to change;” buce- “to die;” ebiàe“to bathe,” “to swim;” benji- “to send (hither),” “to deliver (hither);” buju- “to boil,” “to cook”7; 2) a geminate vowel: wa- “to kill;” sib. /vaam�/ [va�m]; bi- “to be,” “to exist;” sib. /bii/ [bi']; su- “to take off,” “to remove,” “to untie,” “to undo;” sib. /soom�/ [sw���m]; 6
There are not so many homonyms in Manchu, but the pair deheme “aunt” and deheme, the form of the imperfect converb of the verb dehe- “to refine,” “to smelt,” “to temper,” one of them. 7 Phonological and phonetic transcriptions are extracted from the “Classified Dictionary of Spoken Manchu” by Yamamoto.
gorel-1.p65
115
4/17/02, 8:56 AM
116
part three
3) a diphthong: bai- “to ask for,” “to look for;” “to seek,” “to wish;” gai- “to take,” “to be given,” “to get,” “to receive;” sai- “to bite,” “to chew;” goi- “to hit the mark,” “to strike the target,” “to be struck (by an arrow);” nei- “to open,” “to open up;” neo- “to roam,” “to wander away from home;” tu(w)a- “to look,” “to observe,” “to examine,” “to try,” “to visit;” gu(w)e- “to forgive,” “to avoid,” “to escape;” ju(w)e- “to transport,” “to transfer,” “to move;” su(w)aliya“to mix,” “to mix up;” sib. /s�iuliam�/ [������ m]; tu(w)akiya- “to watch,” “to guard;” sib. /tiukiam�/ [t���kiam]; fulmiye- “to bind,” “to tie up,” “to tie together;” oholiyo- “to hold in both hands,” “to take in both hands.” According to Norman, in the variant of the Sibe dialect he has described, the second element of a geminate cluster is omitted before a suffix in a verbal stem (Norman, 1974:168). 6.3. Alternation of Vowels and Consonants in Stems Alternation of phonemes, involving both vowels and consonants in relation to certain lexical and grammatical meanings of words, may be observed in stems. Consequently, in Manchu I encounter alternation of the vowels a and e, which indicates the biological sex in nouns denoting human beings and animals. Since a represents the male or positive principle (yang), it is included in nominal stems conveying masculinity (the notion of masculinity); by contrast, as e indicates the female or negative principle (yin), it is included in nominal stems denoting female meanings (see Part IV, 2.3). Here are some examples: ama “father”—eme “mother;” amha “wife’s father”—emhe “wife’s mother;” haha “male,” “man”—hehe “woman,” “female;” habtaha “a wide girdle used to protect a man’s midsection in battle”—hebtehe “a wide waistband worn by a woman;” amila “the male of animals and birds”—emile “the female of animals and birds;” arsalan “lion”—erselen “lioness;” a i jakdan “pine (which fecundates)”— e i jakdan “pine (which becomes fecundated).” The same alternation can be found in nominal stems where more or less the same semantic opposition can be percieved: ganggan “hard,” “strong”—genggen “soft,” “weak.” This alternation may occur in nominal stems opposed with regard to their meanings or the purpose for which the objects denoted by these stems are used: hadai “plug,” “wedge,” “tap”—hedei “bunghole,” “mortise;” a i bukdan “the outside edge of a piece of folded paper”—e i bukdan “the inside edge of a piece of folded paper;”
gorel-1.p65
116
4/17/02, 8:56 AM
117
phonology
ganggahån “tall”—genggehun “stooped.” This alternation may occur in verbal stems comprising the semantic opposition: wasi- “to descend,” “to go down,” “to fall (of rulers),” “to decline (of value)”—wesi- “to ascend,” “to go up,” “to raise,” “to advance (in rank).” Alternations denoting grammatical meanings are typical for pronouns. Thus, the alternation of b/s/ø shows the difference between personal pronouns with regard to the person which this pronoun indicates: bi “I,” “me”/ si “you” (SG) / i “he,” “she.” The alternation of the vowels i/e indicates the category of number in personal pronouns: bi “I,” “me” / be “we (exclusive);” si “you (SG)”/ se “you (PL).” There are certain vowel alternations in stems which are conditioned by stylistic, dialectal, or historical distinctions. Here are some examples of such discrepancies: alarame /alirame “along a mountain, a low hill;” dahåme/dahime “again;” efin/efiyen/efien “game,” “play;” efi-/efiye-/efie- “to play;” ibiya-/ubiya- “to detest,” “to loathe;” ibiyacun/ubiyacun “loathing,” “disgust;” ifi- /ufi- “to sew;” urile-/urule“to deem right,” “to consider correct;” isibu-/isimbu- “to bring to,” “to deliver,” “to pass to (someone);” gele-/golo- “to fear,” “to be afraid;” ufuhuna-/ufuhune- “to become plump” (Zakharov, 1879:65; Pashkov, 1963:18). 6.4. Vowel Harmony in Suffixes The law of vowel harmony applies to the majority of suffixes denoting lexical or grammatical meanings, with a few exceptions mentioned above (see section 2). Most suffixes change their vowels in accordance with those vowels which occur in a stem, mainly to a vowel of the first syllable of a word. The rules applying to suffixes added to a stem, are the following: 1) If a stem contains the vowel a, suffixes have the same vowel: aca-ha, aca-ra (aca- “to meet,” “to join”); aca-la-ha, aca-la-ra (aca-la“to act together,” “to act mutually”); aca-m-ja-ha, aca-m-ja-ra (aca-mja- “to come together in one place”); aca-na-ha, aca-na-ra (aca-na- “to go to meet,” “to fit,” “to suit”). 2) If a stem contains the vowel o, suffixes have the same vowel: horo-lo-ho, horo-lo-ro (horolo- “to show severity,” “to frighten,” “to inti midate”); obo-ho, obo-ro (obo- “to wash”), obo-no-ho, obo-no-ro (obo-no“to go to wash”).
gorel-1.p65
117
4/17/02, 8:56 AM
118
part three
3) If a stem contains the vowel e, suffixes have the same vowel: bederce-he, bederce-re (bederce- “to retreat,” “to withdraw”); elde-ke, elde re (elde- “to shine,” “to glow”), elde-ne-he, elde-ne-re (elde-ne- “to go to shine”). 4) When a stem contains the vowels i or u (the “soft” u), a more complicated situation arises. Sometimes these vowels must be followed by the vowels a and o in suffixes, sometimes they must be followed by suffixes having the vowel e. Very often one set of suffixes has the vowel a and o after the vowels i and u in stems, but others contain the vowel e: ili-ha, ili-re (ili- “to stand,” “to stand up,” “to stop”); uru he, uru-ke, uru-re, uru-nde-re (uru- “to be hungry,” “to get hungry”); uru ne- “to go about hungry;” uru-le-he, uru-le-re (urule- “to deem right,” “to consider correct”); uruàe-he, uruàe-re (uruàe- “to deem right,” “to consider correct”); uruàa-ha, uruàa-ra (uruàa- “to fulfill the duties of a daughter-in-law”); muri-ha, muri-re (muri- “to twist,” “to wring,” “to pinch,” “to wrong (someone),” “to be stubborn”). According to my data, in classical Manchu, the vowel e occurs in the form of the imperfect participle more frecuently. In the Sibe dialect, this form reveals the tendency towards unalterability. While other suffixes change their vowels under the influence of the vowels of the stem, this form is used with the vowel e: taci-re, gåni-re, but taci-ha and gåni-ha (taci- “to learn,” “to study;” gåni- “to think,” “to reflect”) (Lebedeva & Gorelova, 1994:28). 5) When the first stem vowel is a, and the second one is i, then in some suffixes the vowel a occurs, but in the others the vowel e is required: ali-na-ha, ali-na-ra; ali-ha, ali-re (ali- “to receive,” “to accept,” “to undertake,” “to support,” “to hold up”). 6) If the first stem vowel is o and the second one is i, then in suffixes the vowel a mainly occurs but sometimes the vowel e appears: morila-ha, mori-la-ra (morila- “to ride a horse,” “to go by horse”); olji-laha, olji-la-ra (oljila- “to capture a prisoner during wartime”); oli-ha, oli-re (oli- “to avoid the road,” “to veer to the side,” “to wind”). 7) If the first stem vowel is e and the second one i or u, then the vowel e occurs in all added suffixes: eri-le-he, eri-le-re (erile- “to act at the right time,” “to keep the proper time,” “to do often”); eru-le-he, eru-le-re (eru-, erule- “to torture,” “to punish”).
gorel-1.p65
118
4/17/02, 8:56 AM
119
prolegomena to manchu studies
8) If a stem contains the “soft” vowel u in all syllables, then in suffixes both vowels a and e may occur: ucu-le-he, ucu-le-re (ucu-le- “to sing,” “to mix”); uju-la-ha, uju-la-ra (uju-la- “to head,” “to head up,” “to be in charge”). 9) The same rules apply to diminutive suffixes appended to qualitative nouns (a group of nouns denoting semantics of quality; in most Altaic grammars such nouns are considered to be adjectives). Here are some instances: amba “big,” “great,” “older,” “seniority”— ambakan “rather big,” “a person who is rather large;” labdu “many,” “much,” “many (people)”—labdukan “rather a lot,” “rather many;” onco “wide,” “broad,” “width,” “breadth”— oncokon “rather wide, broad;” sain “good,” “well”—saikan “rather well,” “nicely,” “prop erly;” gulu “simple,” “pure,” “plain”—guluken “rather plain,” “rather unadorned;” idun “coarse,” “rough,” “uneven”—idukan “rather coarse.” 10) The same rules apply to the suffix -ngga/-nggo/-ngge attached to nouns and the suffix -cuka/-cuke (it is used -cuka instead of -cuko) appended to verbs. Here are examples: acan “harmony,””concord,” “union,” “meeting,” “juncture”—acangga “harmonious,” “fitting;” elden “light,” “glory,” “respledence”—eldengge “shining,” “glorious,” “glowing,” “respledent;” horon “authority,” “awe,” “majesty,” “power,” “poison”—horonggo “powerful,” “majestic,” “possessing great authority,” “poisonous;” gånin “thought,” “opinion,” “feeling,” “sense,” “mind,” “spirit”—gåningga “full of ideas,” “reflective;” ulgiyan “swine,” “pig,” “the twelfth of the earth’s branches”—ulgiyangga “pertaining to the pig,” “pertaining to the twelfth cyclical sign;” ulhicun “understanding,” “insight,” “knowledge”—ulhicungga “pos sessing understanding or insight,” ulhicuke “understandable” (ulhi- “to understand,” “to comprehend”); kumun “music”—kumungge “noisy,” “festive,” “exciting,” “lovely,” “animated;” akacun “sadness,” “grief”— akacuka “sad,” “pitiful,” “grievous” (aka- “to be sad,” “to grieve”); gelecuke “frightful” (gele- “to fear”).
gorel-1.p65
119
4/17/02, 8:56 AM
This page intentionally left blank
121
prolegomena to manchu studies
PART FOUR
MORPHOLOGY
gorel-1.p65
121
4/17/02, 8:56 AM
This page intentionally left blank
123
prolegomena to manchu studies
PART FOUR
MORPHOLOGY As mentioned above (see Part I, Section 2) the Manchu language, as compared with other languages of the Altaic language family, is the most analytical one and its morphological apparatus is developed rather weakly. It is difficult to divide all Manchu words into parts of speech because the notion “parts of speech” itself requires the existence of well-developed morphological devices corresponding to certain grammatical functions. Because of weak differentiation of parts of speech, parts of sentence assume greater importance in the organization of the Manchu utterance. All Manchu words can be divided into two numerically unequal groups. They are categorematic (words of full meaning) and syncategorematic words. The categorematic words, in their turn, are distributed into two large classes, morphologically opposed to each other, viz. nouns and verbs. They differ to a great extent with regard to their semantic and morphological characteristics. Verbal markers are so stable and specific that there is no necessity to know the meaning of a word in order to ascribe it to the class of verbs. The suffixes -mbi, -mbumbi, -ka/-ko/-ke, -ha/-ho/-he, -ra/-ro/-re, -habi/-hobi/-hebi, -mbihe, -kini, -me, -fi (-pi), -ci, -cibe and some others unambiguously indicate verbal forms. Noun markers are not so numerous and uniform as verbal ones. Therefore Manchu nouns, as a morphological class, are characterized negatively in the “noun – verb” opposition. There are several classifications of the parts of speech in Manchu. Zakharov divided all Manchu words into the following grammatical classes: 1. nouns (substantives and adjectives), 2. pronouns, 3. verbs (including participles and converbs), 4. adverbs, 5. postpositions, 6. conjunctions, 7. interjections (including onomatopoeic words), 8. particles. Substantives, adjectives, numerals, and pronouns were classified by him as declinable words, verbs were considered as conjucated words (Zakharov, 1878:68-9). According to Möllendorff, all Manchu words can be divided into eight grammatical classes: 1. nouns (substantives and adjectives), 2. pronouns, 3. numerals,
gorel-1.p65
123
4/17/02, 8:56 AM
124
part four
4. verbs, 5. adverbs, 6. postpositions, 7. conjunctions, 8. interjections. Substantives and adjectives were included by him into one grammatical class because of many semantic and morphological similarities between them (Möllendorff, 1892:4). B.K. Pashkov distinguished adjectives, participles, and converbs as separate grammatical classes. According to him, twelve grammatical classes may rightfully be distinguished in classical Manchu. These include substantives, adjectives, pronouns, numerals, verbs, participles, converbs, adverbs, postpositions, conjunctions, particles, and interjections (Pashkov, 1963).
1. Polysemantics and Polyfunctionality of Language Units as Two Principal Features of Manchu Grammar It is well known that it was B. Laufer who considered the polysemantics of some words, particularly verbs, to be the fundamental difficulty of the Manchu language (Laufer,1908). In Manchu there is a very specific class of verbs which is characterized by exceptionally abstract semantics. To this class one can ascribe the verbs tuci-, gene-, ji-, gai-, wa-, yabu-, and some others which cannot be adequately translated into other languages (in this particular case they mean, respectively, “to appear,” “to go,” “to come,” “to take,” “to kill,” “to go,” “to act”), due to the semantics of Manchu lexemes which is considerably broader. It is possible to illustrate this phenomenon, using, as an example, one of the verbs with a broad scope of meanings in Manchu, namely the verbal lexeme tuci- which expresses the abstract concept of appearing, arising, going out (from inside to the outside), irrespective of the specific mode of action. In Zakharov’s Manchu-Russian dictionary the lexical entry tuci contains the following specifications of its general meaning “to appear,” “to come out” (Zakharov, 1875:760-1): 1. Relative to the concept of motion and formation (while changing quality): to appear, to emerge, to come out, to come forth, to arrive; to exit, to leave, to go out, to set off, to drive away, to go on a trip; to get out, to set out; to get married; to become a monk (to take monastic vows).
gorel-1.p65
124
4/17/02, 8:56 AM
125
morphology
2. Relative to vegetation: to grow, to come up, to spring forth, to sprout. 3. Relative to natural phenomena: to rise (of the sun). 4. Relative to a human being and human physiological functions: to be born ( to come out of the womb); to cut (one’s teeth); to appear (about sweat); to break out (about a rash); to belch out, to vomit (something goes out of a human being). The verb stem tuci- may occur with the suffix -bu, which coinsides in form with the passive and causative, and might be the result of their further development. Consequently a wide range of meanings appears, relative to human activities: to take out, to bring out, to remove (from inside something); to take a coffin to the place of burial; to recommend; to reveal, to discover; to appoint, to delegate, to send out (on a mission); to publish; to save, to rescue (NL:282). In order to specify the verbal action expressed by the verb tuci-, as well as by other verbs of this class, these verbal word forms characterized by abstract semantics are used in combination with the form of the imperfect converb (form in -me), derived from verbs having more concrete meaning. Here are some examples: uka-me tucihe “(one) ran away” (literally: uka- “flee,” “ to run away,” tuci- “go out” ); dosi-me gene-fi “(one) entered” (literally: dosi- “to enter,” gene- “to go”); uda-me gai-ha “(one) bought” (literally: uda- “to buy,” gai- “to take”); saci-me wa “slash to death!” (literally: saci- “to cut with a sword,” “hack,” “slash,” wa“to kill”); deye-me yabu-ha “(one) flew away” (literally: deye- “fly,” yabu“to go,” “to leave”). The grammatical structure of the Manchu language is characterized, besides polysemantic verbs, by another specific feature, namely the polyfunctionality of language units, primarily of verbal word forms. This feature serves to compensate for the relatively weak development of Manchu morphology in comparison with the morphological systems of the other languages in this linguistic family. Below we shall demonstrate the polyfunctionality of Manchu word forms using as an example one of the verbal forms, the imperfect converb which is formed with the suffix -me. One of the most com monly used forms in Manchu, it realizes a wide and diverse range of functions (see 5.7.1).
gorel-1.p65
125
4/17/02, 8:56 AM
126
part four
2. Nominal Parts of Speech. Different Functional and Semantic Groups of Nouns For the majority of contemporary specialists it is undisputable that, in all Altaic languages, substantives, adjectives, and adverbs exist as separate grammatical classes of words. S.L. Charekov suggested that adjectives could be morphologically distinguished dated as early as fifth-eighth centuries (Charekov, 1990:118, 121). According to Charekov, by that time special adjective suffixes had already existed as a result of a long term evolution. In his view, the problem of nominal parts of speech has been pertinent only with respect to the so-called syncretic forms, which are formally and semantically identical to nouns having semantics of quality. A theory of nouns of quality appeared for the first time in relation to Mongolian nominal words. Subsequently, a similar theory was also developed as applied to the Tungus-Manchu languages. In addition to the nouns of quality, in Tungus-Manchu languages, it became possible to isolate even such nominal parts of speech as nouns with semantics of place, time, etc. There is another grammatical tradition, according to which substantives and adjectives are considered as being within a single part of speech, viz. nouns. This tradition appeared rather early and stemmed from the grammatical nature of nominal words. Later, from the standpoint of modern linguistics, this way of thinking was considered antiquated, however, the problem of distinguishing nominal parts of speech was not solved. In all Altaic languages there is a particular group of nominal words which denote the meaning of quality in the broad sense. These words are not very numerous, but due to frequency of their use, they are very important in each language. The semantics of quality manifests itself not only through various qualitative characteristics of objects and actions but also through an abstract concept of quality. In a sentence, these nominal words may perform different syntactic roles, e.g., they may be attributes, objects, and adverbial modifiers. Having different functions within the sentence, the nominal words do not normally change their morphological forms. This is why they have been referred to as “syncretic forms.” There is a school of thought which believes that adjectives have, in fact, developed from the syncretic forms. Some of them began to be used in the attributive position directly before the other noun, and later the syntactic role of an attribute was supported by specific suffixes. These, in the course
gorel-1.p65
126
4/17/02, 8:56 AM
127
morphology
of time, became suffixes which were used to mark adjectives as a separate grammatical (morphological) class of nominal words (Charekov, 1990:25-7). It should be emphasized that most nouns of quality are non-derived and archaic. Their archaic character, as considered by many scholars, reflects the stage in the development of the language when contemporary morphological classes of words had not been formed yet (Charekov, 1990:13). Derivative word forms containing pro ductive suffixes can also be found among nouns of quality. They function in the same way as the non-derived syncretic forms, having different roles in the sentence. Occurring in all Altaic languages, these nominal words were specified as adjectives by most scholars. But, in fact, they form a wider class, overstepping the boundaries of adjectives as a grammatical category. Realizing this, many specialists attempted to create a conceptual framework in which the nouns of quality would find their proper place among other nominal words. A. Bobrovnikov was the first to replace the traditional terms “substantives” and “adjectives,” which were predominantly used in Mongol studies before him, with the following terms: object nouns, nouns of quality and relative nouns. The nouns of quality were defined by him as names of quality intrinsic to an object: sayin “good,” “kind,” “the good,” “kindness;” ünen “true,” “truth;” ma�u “bad,” “evil,” “the bad;” qar-a “black,” “blackness;” narin “thin,” “thinness.” However, he emphasized that there are no clear boundaries between nouns with semantics of quality and those with semantics of object (Bobrovnikov, 1849:525). G.D. Sanzheev postulated four classes of nominal words, viz. substantives, object nouns, adjectives, and nouns denoting semantics of quality. According to him, nouns of quality denote various qualitative characterictics which are ascribed to substantives. They perform two syntactic functions in a sentence, serving as attributes and adverbial modifiers of manner: sayin “good,” “well;” ma�u “bad,” “badly;” qurdun “quick,” “rapid,” “fast,” “quickly,” “rapidly” (Sanz heev, 1953:124-6). Although most contemporary specialists categorize adjectives as a separate grammatical class in Mongolian, they note that there are no universal morphological suffixes in their stems according to which it would be possible to classify nominal words as adjectives (Nadelyaev, 1988:96-7). There are morphological markers, some of them being very productive, but they can be found only in the sphere of derivation (Orlovskaya, 1961:84-94). In addition,
gorel-1.p65
127
4/17/02, 8:56 AM
128
part four
qualitative adjectives have degrees of quality as well as forms denoting the intensity of quality. All these forms are considered as grammatical characteristics constituting Mongolian adjectives (Orlovskaya, 1961:95-9). Adjectivity, as the category semantics of adjectives, can manifest itself indirectly, through the combination of adjectives with substantives. Normally, serving as attributes, adjectives occur before substantives; while acting as predicates, they take place after substantives. V.M. Nadelyaev devised a special test, which makes it possible to know whether or not a nominal word is an adjective. This test can be carried out through the verification for predicativity. If a nominal word may serve as a predicate, then it can be classed as an adjective (Nadeliayev, 1988:97-8). Specialists in Turkic languages, note that nominal words, which semantically correspond to adjectives in some other languages, represent a very mixed group of words, highly specific to each Turkic language. Denoting various qualitative characteristics of objects, these words do not have suffixes which express the semantics of adjectives as a grammatical category. Normally they serve as attributes and predicates. When functioning as attributes, they stand directly before substantives. Serving as predicates, they follow the substantives. Nominal words with semantics of quality are very limited in numbers. These words denote not only qualitative chatasterictics of objects, but quality as such, i.e. as an abstract notion. For example, in the Tofalar language one can find the following words which have semantics of quality: bedic “high,” “height;” cilig “warm,” “warmth,” “heat;” hilin “thick,” “thickness;” ol “wet,” “humidity,” “moisture;” sin “right,” “true,” “correct,” “truth;” uzun “long,” “length” (Rassa din, 1978:84). There are some derivative subgroups of nominal words which are semantically associated with adjectives. But some of the derivative suffixes are used to form both substantives and adjectives, and the others both adjectives and adverbs. They have degrees of quality that can be considered as an important argument for classifying them as adjectives. Several turkologists define substantives, which occur in possessive forms, as adjectives proper. According to E.I. Ubryatova, in Yakut, the suffixes semantically associated with the category of possession, are rather adjectival than possessive (Ubryatova, 1950; 1976). In the parent Turkic language, categorial semantics of adjectives has no morphological expression, and the existence of adjectives, as
gorel-1.p65
128
4/17/02, 8:56 AM
129
morphology
a grammatical class of words, seems to be questionable (Scherbak, 1977:108). In contemporary Turkic languages, adjectives show a tendency toward their differentiation from other nominal words. But the process itself has not been completed yet. Nominal parts of speech are very problematic in all Tungusic languages, especially in Manchu. Traditional division into substantives and adjectives does not correctly explain the real situation in the language. That is why, since the very inception of Tungus-Manchu studies till now, several classifications of nominal words have been suggested. G.M. Vasilevich believed that besides substantives and adjectives, there were some particular grammatical classes of nominal words in Evenki. Words, which can be ascribed to these classes, denote semantics of quality, quality and state, emotion and state, quantity, space and time, etc. (Vasilevich, 1958:689, 702). Belonging to the same grammatical class, nominal words normally perform different syntactic functions. Accordingly, they may correspond to different parts of speech in other languages and, therefore, can be translated into these languages by different parts of speech. All of them have full or partial word formation, which is typical for nouns. In the Evenki language the nouns of quality may denote quality as an attribute modifying an object (serving as adjectives), as an attribute modifying an action (serving as adverbs), and as an abstract notion (serving as substantives). Most of them combine the category semantics of both substantives and adjectives, the others represent non-differentiated forms of both adjectives and adverbs. There are words which combine the categorial semantics of substantives, adjectives and adverbs, not differentiating among them formally. The following nominal words have semantics of quality: ala “tasty,” “taste;” albin “wide,” “width;” erå “ bad,” “harm;” gugda “high,” “height;” delum “secret,” “a secret;” haktira “dark,” “darkness;” hegdi “large,” “big,” “magnitude,” “size;” kiri “dirty,” “dirt;” nekte “low,” “a low place;” �¿rÊ “light,” “the light;” �onim “long,” “lenght;” nyama “warm,” warmth,” “warmly;” su�ta “deep,” “depth;” ul�k “lying,” “mendacious,” “the lie”1. Here are some examples where the word aya “good,” “kind,” “the good,” “kindness,” “well” occurs in different syntactic functions: as 1
gorel-1.p65
The bar here is used to indicate the length of vowels in the Evenki language.
129
4/17/02, 8:56 AM
130
part four
an attribute, an object, an adverbial modifier of manner, or a predicate: aya beye “good man,” “kind man” (the word aya is used as an attribute); aya-va-n sa�-de-m the good-ACC-3.SG.POSS know-ASP-1.SG.PR(PERF)
“I know about his kindness” (the word aya functions as an object);
aya-t duku-m
good-INST write-1.SG.PR(PERF)
“I have written (smth.) well” (the word aya is used as an adverbial
modifier of manner);
aya-ka�kun duku-m good-INTSF write-1.SG.PR(PERF)
“I have written (smth.) very well” (the word aya is followed by the
intensifier ka�kun);
tare asi aya-ka�kun (bisi-n)
that woman good-INTSF COP-3.SG.PRS
“That woman is very good” (the word aya is followed by the intensifier
ka�kun; it functions as a predicate).
In Evenki, nouns of quality may be followed by case forms and suffixes of plurality. They also may be appended with personalpossessive suffixes. They have degrees of quality. The following examples show how personal-possessive suffixes are attached to the nouns of quality: bira su� ta-n river depth-3SG.POSS “the depth of a river;” ure-l gugda-tin mountain-PL height-3.PL.POSS “the height of mountains;” gugda-v height-1.SG.POSS “my height” (Vasilevich, 1958:702). O.A. Konstantinova considered nominal words, which have semantics of quality and are mostly syncretic forms, as qualitative adjectives. The semantics of these nouns was the main reason for classifying them as adjectives. She noted that these nominal words denote quality as abstract notion and therefore can be used as substantives. They also modify verbs and therefore can be used as predicative adverbs
gorel-1.p65
130
4/17/02, 8:56 AM
131
morphology
(Konstantinova, 1964:103). She gave some examples of use of such nominal words: albin bira wide river “a wide river” (the word albin is used as an adjective); bira albin-i-n umuken� kilometra river width-CONN-3.SG.POSS one kilometre “The width of the river is one kilometre” (the word albin is used as an object); erå tirganÊ bad day
“a bad day” (the word erå is used as an adjective);
sap erå-ve-n gåne-m
tobacco harm-ACC-3.SG.POSS speak-1.SG.PR(PERF)
“I spoke on the harm of tobacco” (the word erå “harm” is used as
a substantive).
Besides traditional parts of speech in the Nanai (Gold) language, such as substantives, adjectives, and numerals, V.A. Avrorin proposed the establishment of several separate grammatical classes of nominal words, such as nouns with semantics of quality, nouns with semantics of time, and nouns with semantics of negation (Avrorin, 1959:2229). According to his theory, nouns of quality and nouns of time are related to adverbs which form a separate grammatical class of words (Avrorin, 1959:103-4). He also noted that nouns of quality may be used as objects, attributes, adverbial modifiers and predicates (Avrorin, 1968:136-7). Here are some examples: ulen si�akta good fur “good fur” (the word ulen is used as an attribute); si�akta ulen fur good “Fur is good” (the word ulen is used as a predicate); ulen hola-j-ni good read-PRS-3.SG.PR “(One) reads well” (the word ulen is used as an adverbial modifier of manner); ulem-be ba-ha-ni good-ACC get-PAST-3.SG.PR “(One) has got a pleasure (the word ulen is used as an object).”
gorel-1.p65
131
4/17/02, 8:56 AM
132
part four
M.M. Khasanova, in the article written on the occasion of Avrorin’s ninetieth birthday, mentions Avrorin’s proposal, found in his unfinished and unpublished description of classical Manchu (now published in Avrorin 2000), to establish a separate grammatical class of nominal words which would be called nouns of quality. According to him, in Manchu there is a large and very important grammatical class of nominal words which denote semantics of quality in the broad sense. Semantically and grammatically these nominal words correspond to adjectives, numerals, adverbs and partly to substantives in some other languages (Khasanova, 1998:109). In the description of Sibe (a language spoken by the Manchu tribe Sibe), instead of following the traditional division of nominal words into such grammatical classes as substantives and adjectives, E.P. Lebedeva proposed a division based on different functional and semantic classes of nominal words (Lebedeva & Gorelova, 1994). This view, which I mostly support, is presented below. In the system of Manchu nominal words the isolation of several large classes of words which are comparable with nominal parts of speech in other languages appears justifiable. As mentioned above, the notion “part of speech” itself presupposes the existence of a set of morphological markers, characterizing a certain semantic class of words. Since in Manchu there is no nominal class having its own morphological markers, typical for only this class of words, one can rightfully speak about various functional and semantic classes within a single part of speech, i.e. the noun, rather than about different nominal parts of speech. In the Manchu language there is only one morphological category common for all groups of nouns, that is declension. But declension can not be considered as a special category unique to the nominal class, or even to all nouns. In Manchu, this grammatical category is the universal system which conveys syntactic subordination and extends to nouns, pronouns, participles, syntactic words, and clauses as well. Opposition of singulars and plurals, expressed by special plural suffixes, is generally limited to nouns denoting human beings (human nouns) and pronouns. Other nominal classes do not have the morphological category of plurality at all, i.e. plural suffixes cannot be attached to them. Morphological markers, typical for certain semantic classes of nouns, may be found only in the sphere of word formation. Normally
gorel-1.p65
132
4/17/02, 8:56 AM
133
morphology
a certain nominal class has its own specific derivational suffixes. However, due to the fact that every class includes both basic (non derived) nouns, which have no derivational suffixes at all, and borrowed words, it becomes impossible to ascribe nouns to certain nominal classes on the grounds of sharing a particular inventory of derivational morphemes. Functions of nouns in a sentence cannot be considered as the right criterion for differentiating various nominal classes since all nouns may serve as any part of a sentence. Thus, the only criterion according to which it is possible to divide all Manchu nouns into classes, is the semantic one. In accordance with lexical and semantic characteristics, all Manchu nouns should be ascribed to the following classes: 1) nouns denoting human beings (human nouns), 2) nouns denoting inanimate objects and abstract notions (object nouns), 3) nouns having semantics of quality that is understood either as an attribute to a noun or as an abstract concept. Very often nouns of this type denote a human being, characterized in accordance with the qualitative feature he/she possesses. The term “substantive” normally corresponds to the group of nouns having the semantics of the object in the broad sense. These nouns are characterized by common morphological and syntactic features according to which the group may be classified as a separate morphological class. In the Manchu language we can speak about two groups of nouns which could be brought into correlation with substantives. One of them includes nouns denoting human beings, the other comprises nouns denoting inanimate objects and abstract notions. They are called human nouns and object nouns. 2.1. Human Nouns The group of human nouns comprises all names denoting human beings according to their age and biological sex, post and rank, line of business, ties of relationship and proper names. From the point of view of their semantics, these words are opposed to the rest of nouns. Human nouns may be classified not only based on semantics, but also based on their morphological characteristics. Thus, only human nouns may correlate with the pronoun we “who?;” other nouns can be brought into correlation with the pronoun ai “who?,” “what?,”
gorel-1.p65
133
4/17/02, 8:56 AM
134
part four
“which?”. The pair combination we ai “who?” (about a number of persons) is the plural form for the interrogative pronoun, correlating with human nouns. Human nouns, in contrast to the rest, can be antecedents of the personal pronouns i “he,” “she” and ce “they,” used anaphorically. The rest of the nouns can be antecedents of the demonstrative pronouns ere “this” and tere “that,” used anaphorically. The demonstrative pronouns can also be used anaphorically to replace human nouns, competing in this role with the personal pronouns. Human nouns, unlike other groups of nouns, have the morphological category of number. It may rightfully be said that plurals are generally limited to human nouns and pronouns. 2.1.1. Plurals The most commonly used suffixes for plurals (or more properly collectives) are -sa/-se/-so, -ta/-te, -si and -ri. The suffixes -sa and -se are used with nouns denoting age, generation and relatives (kinsmen and kinswomen): ecike “father’s young brother (uncle)”—ecikese “father’s young brothers (uncles);” gege “elder sister” (also a respectful term of address to young ladies)— gegese “elder sisters;” jui “son,” “child”—juse “sons,” “children;” sadun “father of the son-in-law,” “father of the daughter-in-law”—sadusa “fathers of the son-in-law, the daughter-in-law;” sagda “old man”— sagdasa “old men”2. These suffixes are also used with nouns denoting peoples and nations, posts, ranks, titles and occupations: age “prince, son of an emperor” (also a polite term of address, master, sir, lord)—agese “princes, sons of an emperor” (also sirs, masters); amban “high official”—ambasa “high officials,” antaha “guest”—antahasa “guests;” bayan “rich man”—bayasa “rich men;” beile “ruler,” “prince of the third rank”—beile se “rulers,” “princes;” faksi “craftman,” “workman”—faksisa “craftmen,” “workmen;” gucu “friend”—gucuse “friends;” hafan “official,” “officer”—hafasa “officials,” “officers;” han “khan,” “emperor”—han sa “khans,” “emperors;” irgen “people,” “nation”—irgese “peoples,” “nations;” lama “lama,” “monk”—lamasa
2 Norman translates the word sadun as “related by marriage,” “a relative by marriage” (NL:230).
gorel-1.p65
134
4/17/02, 8:56 AM
135
morphology
“lamas,” “monks;” manju “Manchu”—manjusa “the Manchus;” nikan “Chinese”—nikasa “the Chinese;” oros “Russian”—oros se “the Russians;” àabi “pupil,” “student”—àabisa “pupils,” “students”3. The suffix -so is found in the names of peoples and nations: monggo “Mongol”—monggoso “the Mongols;” solho “Korean” – solhoso “the Koreans;” also gioro “the name of the Manchu dynasty clan”—gioro so (in old Manchu books it was often used in the form gioro se); giohoto “beggar”—giohoto so “beggars.” The suffix -si is used with some nouns denoting age, relative relationships and status of human beings: aha “slave”—ahasi “slaves;” haha “male,” “man”—hahasi “males,” “men” (the polite plural form haha niyalma); hehe “female,” “woman”—hehesi “females,” “women” (the polite plural form hehe urse); hojihon “son-in-law”—hojihosi “sonsin-law;” omolo “grandson”—omosi “grandsons.” The suffixes -ta and -te are used with words denoting the age of human beings, generation and relatives: ahån “elder brother”—ahåta “elder brothers;” ama “father”—amata “fathers;” ambu “mother’s elder sister”—ambuta “mother’s elder sisters;” amji “father’s elder brother”—amjita “father’s elder brothers;” amu “father’s elder sister”—amuta “father’s elder sisters;” asihan “young man”—asihata “young men;” aàa “elder brother’s wife”—aàata “elder brother’s wives;” efu “the husband of one’s elder sister,” “wife’s elder brother,” “the husband of wife’s elder sister” – efute “husbands of one’s elder sister,” “wife’s elder brothers,” “husbands of wife’s elder sister;” eigen “husband”—eigete “husbands;” ejen “ruler,” “lord,” “master,” “emperor”—ejete “rulers,” “lords,” “masters;” eme “mother”—emete “mothers;” eshen “father’s younger brother (uncle)”—eshete “father’s younger brothers (uncles);” eyun “elder sister”—eyute “elder sisters;” da “foreman”—data “foremen;” dehema “husband of mother’s sister (uncle)”—dehemata “husbands of mother’s sister (uncles); deheme
3 The plural suffix -se should be distinguished from the morpheme -se which some Chinese loan words contain. In the Chinese borrowings the morpheme -se (in Manchu transcription) represents the Chinese morpheme -zi: ma. cise “pond” < chin. chizi; ma. dangse “register” < chin. dangzi; dengse “steelyard” < chin. dengzi; kose “pants,” “trousers” < chin. kuzi; pase “a rake” < chin. pazi; panse “plate,” “tray” < chin. panzi; tangse “the imperial shamanic shrines in Peking and Mukden” < chin. tangzi ( Zakharov, 1879:121; CD). Norman translates the word panse as “chess board,” “measure for games of chess” (NL:226).
gorel-1.p65
135
4/17/02, 8:56 AM
136
part four
“mother’s younger sister (aunt)”—dehemete “mother’s younger sisters (aunts);” deo “younger brother”—deote “younger brothers;” ge “husband’s elder brother”—gete “husband’s elder brothers;” gu “father’s sister”—gute “father’s sisters;” gufu “husband of father’s sister”—gufute “husbands of father’s sister;” meye “younger sister’s husband “—meyete “younger sister’s husbands” (in old language this word also meant “wife’s younger brother”—NL:197); nakcu “mother’s brother”—nakcuta “mother’s brothers” (also nakcusa, nakcuse); naca “wife’s elder brother”—nacata “wife’s elder brothers;” non “younger sister”—nota “younger sisters;” oke “the wife of father’s younger brother”—okete “wives of father’s younger brother;” sargan “wife”— sargata “wives” (also sargasa); uhume “father’s younger brother’s wife” – uhumete “father’s younger brother’s wives;” ungga “the elder in generation”—unggata “elders in generation” (Zakharov, 1879:120-3; Pashkov, 1963:19-20). Suffix -ri occurs only with a few words: mafa “grandfather”—mafari “grandfathers,” mama “grandmother”—mamari “grandmothers.” Both words are also used in the meaning “ancestors,” “forefathers.” One can see that some words are used with various plural suffixes: agu = a respectful term of address for men: sir, master—agusa, aguse; nakcu “uncle (mother’s brother)”—nakcusa, nakcuse, nakcuta “uncles (mother’s brothers);” sargan “wife”—sargasa, sargata “wives;” urun “daughter-in-law,” “sister-in-law”—urusa, uruse “daughters-in-law,” “sisters-in-law.” There is no strict rule governing whether the plural suffix should be written together with a noun stem or separately (for instance: beile se, han sa, oros se). But in most cases the plural suffix and a noun stem are written as one word. It should be noted that final -n, -i and -lo are deleted in some words before the addition of the plural suffix: amban “high official,” “dignitary”—ambasa “high officials;” jui “son,” “child”—juse “sons,” “children;” omolo “grandson”—omosi “grandsons.” According to V.I. Tzintzius, the Manchu suffixes -sa/-se/-so, -ta/ -te, -si, -ri correspond with the following composite suffixes in the Tungusic languages: evenk., even., neg., ulch., nan. -sal/-sel/-sol < -sa/-se/-so + l; evenk., even., neg. -tal, -til < -ta + l, -ti + l; ud. -nta/ -nte/-nto < n + -ta/-te/-to; evenk., nan. -ril < -ri + l; even. -rel < -re + l; ulch. -ril < -ri + l, -rul < -ru + l. Only the suffix -l occurring alone or in combinations with other suffixes is considered to be a universal marker of plurality. The other components of these
gorel-1.p65
136
4/17/02, 8:56 AM
137
morphology
composite suffixes coincide with those which can be found in collective nouns, numerals, and other parts of speech in Tungusic. In collective nouns, these suffixes convey the meaning of various groups of people or collections of objects which are the results of human activities. In plurals formed by the composite suffixes in which a morpheme for collectivity and a morpheme for plurality have merged, the idea concerning human society, the tribal system, clans and kindred groups of people becomes more emphasized. The archaic character of this group of nouns, its isolation from the others and, more importantly, the fact that it denotes the collective and comitative meanings, all leads to the conclusion that historically a grammatical category conveying the collective meaning preceded the grammatical category of plurality in Tungusic. As far as Manchu is concerned, it is difficult to ascertain whether the language had lost the suffix -l denoting plurality in Tungusic, with the exception of Udeghe and Oroch, or had never had it (Tzintzius, 1946:73-119). In Manchu, the suffix -sa/-se had the collective meaning in the past. The moribund collective morpheme -se can be found in the Manchu word urse “people.” This suffix corresponds to the Ude word se “clan (a group of persons originating from a male ancestor and united by recognition of their blood relationship),” “nationality,” “family,” “kind,” “sort,” “type.” In its turn the Ude word se is a part of a semantic group of words related to the word “blood” in Tungusic: evenk. sekse, neg. sakse, oroch. saksa, ud. sake, ulch. sekse, nan. sekse, ma. senggi; also neg. senggi “relatives by marriage;” ma. senggi o- “to drink blood,” senggile- “ to act in a bloody manner;” senggileme afa- “to fight a bloody battle;” senggime “bloody,” “love between brothers;” “kindred love;” “intimite,” “friendly,” “on good terms.” The Manchu suffix -ta/-te corresponds to various pronominal elements in Tungusic: the Manchu-Tungus demonstrative pronoun ta-/te- “that,” personal pronoun for the third person, plural ta-/te“they,” possessive suffix for the third person, plural -tan/-ten in Even (Lamut) and -tin in Evenki. It is more difficult to determine the origin of the suffix -ri since there are no words beginning with r in the Manchu-Tungus languages (and in general, in Altaic). It probably corresponds to the same component found in the Tungusic demonstrative pronouns: even., evenk. e-r, e-ri, e-re “this;” ta-r, ta-ri, ta-ra “that.” This suffix is not widely spread.
gorel-1.p65
137
4/17/02, 8:56 AM
138
part four
Tzintzius came to the conclusion that grammatical categories denoting collective and comitative meanings preceded the grammatical category of plurality in the Tungus-Manchu languages. These categories originated from pronouns, collective nouns, and names of human groups, that is nation, people, clan, tribe (Tzintzius, 1946:73-119). In Manchu, the plural may be formed analytically, by adding words denoting collective meaning or plurality. Some of these words like eiten “all,” “every;” geren “all,” “many,” “numerous,” “the various …;” yooni “all,” “all together,” “complete,” “entire;” labdu “many,” “much;” ududu “some,” “several,” “many,” “a number of,” normally precede nouns. Here are some examples: eiten jaka “everything,” “every object;” geren niyalma “all/many men/persons.” The word geren is also used in combination with plural forms: geren ambasa hafasa “many officials;” geren àabisa “many pupils,” “many students.” This is a single case when synthetic means of expressing plurals combine with an analytical one (pleonasm). Other words having the same semantics, such as gemu “all,” tome “every,” each,” always follow nouns: bayan gemu “all rich men;” gurgu tome “ every wild animal;” irgen gemu “all people;” moro tome “every bowl;” niyalma tome “every man/person,” “people;” yadahån gemu “all poor men” (Pashkov, 1963:20). Plurality may be expressed by adding generic words denoting the biological class of living beings: gasha “bird,” gurgu “wild animal,” “beast;” niyalma “man,” “person;” ulha “domestic animal,” urse “men,” “people,” “persons.” The word urse chiefly follows nouns of quality or participles: bayan urse “rich men;” ehe urse “the villains;” tacire urse “the scholars” (taci-re = taci- “to learn” , “to study” + -re , the suffix for the imperfect participle). The words hacin “kind,” “sort,” “class,” “item” and jergi “class,” “degree,” “group,” “rank” should be included in this group. All these words are placed after the relevant nouns. Plurality can be expressed with the help of the words hacin “kind,” “sort,” “class,” “item” and jergi “category,” “grade,” “rank,” “sort.” Following the marker for the genitive case added to the noun, these words denote both plurality of objects and the belonging of the latter to a certain class, kind or sort. Following the noun directly, they denote plurality with the shade of meaning “of different/various kinds, every kind.” The word jergi is normally used with animate objects and the word hacin with inanimate objects. Here are some examples: baita hacin “different/various matters/affairs;” giyahån jergi
gorel-1.p65
138
4/17/02, 8:56 AM
139
morphology
gasha “falcons, hawks and other related birds;” gurgu jergi “different/ various wild animals,” “animal world;” kooli hacin “regulations and precedents;” orho hacin “herbs,” “plants;” tasha jergi gurgu “the tigers and other related beasts;” tasha i jergi gurgu “beasts belonging to the tiger family;” tubihe hacin “fruits” (Zakharov, 1879:123). In Manchu, special combinations consisting of two words are used to denote the collective meaning. The first word is represented by a cardinal number which is an attribute to the following noun. Such combinations originated from the Chinese patterns: duin ergi “the four corners of the world” (literally, “the four sides” < chin. sifang “(four) all sides,” “(four) all quarters”); duin mederi “the whole country,” “all kingdoms,” “the whole world” (literally, “the four seas washing the earth” < chin. sihai “the four seas,” “the whole country,” “the whole world” ); sunja hacin i jeku “cereals” (literally, “five kinds of cereal” < chin. wugu “the five cereals: rice, two kinds of millet, wheat and beans”); tanggå hafan “the officials” (literally, “one hundred officials” < chin. baiguan “one hundred officials,” “the officials” ); tanggå hala “people” (literally,”one hundred families” < chin. baixing “one hundred family names,” “common people”); tumen baita “a great number of affairs” (literally, “ten thousand affairs” < chin. bainshi “a hundred affairs,” “all kinds of affairs,” “numerous affairs”); tumen jaka “all things,” “all creations,” “nature” (literally, “ten thousand things” < chin. wanwu “ten thousand things,” “all things on the earth,” “all nature”) (Zakharov, 1879:124; CD). The plural of nouns can be formed by repeating the noun; i.e. reduplication: jalan “generation,” “world”—jalan jalan “generations,” “worlds;” se “age,” “year (of age)”—se se “ages,” “years.” The combination of two synonyms can also serve to indicate plurality of nouns: baita, sita “matter,” “affair”—baita sita “matters,” “affairs;” gasha, cecike “bird”—gasha cecike “birds;” jergi, hacin “class,” “kind,” “sort”—jergi hacin “classes,” “kinds,” “sorts;” ulin, nadan “goods,” “property,” “possessions,” “wealth”—ulin nadan “riches,” “goods,” “possessions;” ulha, ujime “domestic animal”—ulha ujime “domestic animals” (Zakharov, 1879:123-4; Pashkov, 1963:20). The above-mentioned analytical means of plural formation are analogous with those in the Mongol and Chinese languages. When it is necessary to express the plurality of a number of nouns, enumerated one after another, the plural formatives sa/se, ta/te are used, but they only follow the last noun:
gorel-1.p65
139
4/17/02, 8:57 AM
140
part four
beile beise gung sa-i jergi sira-ra de PL-GEN rank inherit-PART DAT beile “prince of the third rank,” beise “prince of the fourth rank,” gung “duke” (NL:27, 114); “When inheriting various ranks such as beile, beise, gung … ” (Zakh:124). 2.2. Nouns Denoting Inanimate Objects and Abstract Notions (Object Nouns) Object nouns, used to refer to objects in the broad sense, represent the numerous group of words most of which are names for inanimate objects. In this group of basic nouns we can include nouns denoting weapons: beri “a bow,” gida “lance,” “spear,” loho “sword,” niru “arrow;” buildings and their parts: boo “house,” fa “window,” fu “wall,” nahan “underfloor heating flue,” uce “door,” “doorway,” “entrance,” yamun “the court,” “palace,” “a government office;” utensils and tools: agåra, ahåra “tool,” “utensil,” anakå “key,” buleku “mirror,” “looking glass,” coo “shovel, “spade,” fengseku “a small porcelain pan or bowl,” fila “dish,” “plate,” fulhå “bag,” “sack,” futa “rope,” gohon “hook,” suhe “ax,” “hammer;” materials and matters: aisin “gold,” sele “iron,” menggun “silver,” wehe “stone;” plants, their parts, and their groups (clumps): abdaha “leaf,” bujan “forest,” gargan “branch,” hailan “elm tree,” ilha “flower,” jakdan “pine,” moo “tree,” “wood,” “stick”, “pole,” orho “grass,” use “seed;” landscape and its components: alin “mountain,” bigan “field,” bira “river,” ekcin “bank (of a river),” “shore,” hada “mountain peak,” mederi “sea,” “ocean,” muke “water,” “river,” “stream,” na “earth,” “ground,” omo “lake,” tala “plains,” “steppe,” “flatland,” ula “large river;” natural phenomena: aga “rain,” akjan “thunder,” edun “wind,” juhe “ice,” nimanggi “snow,” talman “fog,” “mist,” usiha “star;” nouns with time semantics: aniya “year,” bolori “autumn,” coro “the day after tomorrow,” erin “time,” forgon “season,” juwari “summer,” niyengniyeri “spring,” sikse “yesterday,” tuweri “winter;” nouns having part-of-day semantics: dobori “night,” erder “early in the morning,” inenggi “day,” yamji “evening;” nouns having place semantics: amargi “back, “north,” ba “place, “locality,” “spot,” dergi “top,” “head,” “east,” dorgi “the inner part,” dulimba “middle, “ “center,” fejergi “bottom,” siden “interval space;” animals: buhå “deer,” coko “chiken,” damin “eagle,”
gorel-1.p65
140
4/17/02, 8:57 AM
141
morphology
gasha “large bird,” geo “mare,” gålmahån “rabbit,” “hare,” indahån “dog,” honin “sheep,” losa “mule,” morin “horse,” niman “goat,” singgeri “mouse,” “rat,” tasha “tiger,” ulha “domestic animal,” yarha “leopard;” the human body and its parts: angga “mouth,” femen “lips,” ilenggu “tongue,” oforo “nose,” yasa “eye,” àan “ear,” fehi “brain,” “brains,” cira “face,” uju “head,” gala “hand,” bethe “leg,” meifen “neck,” meiren “shoulder,” fahån “liver,” niyaman “heart,” silhi “gall bladder;” clothing: etuku “clothing,” “garment,” mahala “hat,” “cap,” fakåri “Manchu style trousers,” umiyesun “belt,” “sash,” gålha “boots,” sabu “Manchu shoes,” wase “socks,” tohon “button,” wadan “poket;” food and drink: bele “rice,” buda “meal,” “cooked cereals,” cooked rice,” hibsu “honey,” jeku “grain,” jemengge “food,” “cereals,” ira “millet,” sile “meat soop,” “broth,” ufa “flour,” “meal,” yali “meat.” In a certain sense, nouns denoting abstract notions of various kind, viz. feelings, relations, states, actions, can also be ascribed to this groups of nouns. Here are some examples of nouns denoting abstract notions: boljon “agreement,” cihalan “desire,” “wish,” girucun “shame,” “disgrace,” jobolon “harm,” “trouble,” “disaster,” “calamity;” mujin “ambition,” “aim,” “will,” sebjen “joy,” “gladness,” turgun “reason,” “motive,” “circumstances;” urgun “joy,” “felicity,” “happiness.” None of the object nouns are followed by the plural suffixes. The plurals are often expressed analytically, by adding numerals or words denoting quantity, plurality and collective meaning. Here are some examples: gemu jaka “all things,” nadan inggeni “seven days,” ninggun moo “six trees,” uheri ulin “the whole property,” “all possessions.” Plurality can be expressed by reduplication or using a pair of synonymous words: hutu ibagan “devils” (hutu “devil,” “ghost,” “dis embodied spirit” and ibagan “monster,” “apparition,” “phantom”); etuku adu “clothing” (etuku “clothing,” “clothes,” “garment” and adu “garment”). Object nouns have the category of declension, more detailed description of which will be given in sections devoted to the noun and participle declension (see 2.8 and 5.6.2). Nouns of this group may serve to indicate any part of a sentence. Occurring as one of the secondary parts of a sentence, object nouns normally take certain case markers. But in certain circumstances, when the semantic context allows them to act in such a manner, object nouns may occur in the form of their stems, without markers of cases. These nouns may denote not only an object as such, but may also
gorel-1.p65
141
4/17/02, 8:57 AM
142
part four
characterize an object with respect to its quality. Due to this ability they may serve as attributes in a sentence. Acting as attributes, these nouns mostly denote materials of different kinds: aisin “gold”—aisin weihe “gold horn;” senggi “blood”—senggi jugån “blood road,” “blood vessel;” sun “milk”—sun nimenggi “milk butter.” In the ancient period of Manchu, the tendency of these nouns to function as attributes can be seen even more clearly. There is a considerable number of composite words developed from those stable noun-combinations, in which the first noun is an attribute to the second noun. Some instances are as follows: boo nimaha “whale” was formed by two nouns: boo “house” + nimaha “fish,” where the noun boo “house” functions as an attribute to the noun nimaha “fish” on the syntactic level. On the semantic level only the component “big,” which is included in the semantic structure of the noun boo “house,” is activated; the quality “big” normally associated with houses. The composite word ulme faksi “a workman who makes needles” is derived from the combination the first word of which is ulme “needle” acting as the attribute to the second word faksi “workman” (Zakharov, 1875:162, 509; Lebedeva & Gorelova, 1994: 36-7). When object nouns function as attributes, the marker of the genitive case i is frequently attached to them as in: boo i fu “wall of a house,” daifu i sargan “a doctor’s wife,” nimeku i turgun “a cause of the illness,” gurun i ejen “the khan of a state.” The genitive marker singular i is obligatory when a noun is used to indicate an attribute occurring with several homogeneous parts of the sentence (parts of the sentence which perform the same syntactic role and refer to the same word in a sentence): hooàan’i etuku mahala morin etufi yalufi “(someone) put on paper clothes, hats and mounted the paper horse.” The attribute hooàan’i “made from paper” applies equally to three nouns, namely etuku “clothes,” mahala “hat” and morin “horse.” Nouns with place and time semantics form two separate subgroups within the object nouns class. Nouns of place include nouns related to orientation in space: dergi “top,” “head,” “east,” “upper” “above,” “over;” fejergi “bottom,” “underneath;” dorgi “the inner part,” “inside;” tulergi “the outer part,” “outside;” amargi “back,” “north,” “behind;” julergi “front,” “south,” “in front of;” cargi “the opposite side,” “that side,” “opposite;” wargi “west,” “right (side),” “under,” “underneath;” hanci “nearness,” “closeness,” “near,” “close;” goro “distance,” “distant,” “far,” “far away,” “far off,” “far (from),” “a long way off.” These nouns may function in a sentence either as subject and object,
gorel-1.p65
142
4/25/02, 12:23 PM
143
morphology
or attribute and adverbial modifier of place. Below some instances of nouns illustrating the semantics of place are given: dergi ergi de emu sourin (ma. soorin) bi east side DAT one throne COP “There is a throne on the east side of the house;” dorgi de muduri meihe umuài (ma. umesi) labdu inside DAT dragon snake very many “There are a great number of dragons and snakes inside;” ere amargi ba-de emu amba daimin bi this north place-DAT one large eagle COP “There is a big eagle in the north of that settlement” (SK). Nouns denoting various notions related to time (seasons, parts of day, etc.) can be ascribed to object nouns: inenggi “day;” enenggi “today;” dobori “night,” “at night;” yamji “evening,” “in the evening,” cimari “tomorrow,” “morning,” “tomorrow morning;” niyengniyeri “spring,” “in spring.” Nouns with time semantics may function in a sentence as a subject, an object, attribute, and adverbial modifier of time. In these roles they have no markers attached. When serving as an adverbial modifier of time, these nouns, in the form of a stem, may act as functional analogue of adverbs of time. As one can see, there are no morphological devices to distinguish adverbial and nominal (subject, object, attribute) functions of these nouns. The substantive function can only be distinguished from the adverbial one within the syntactic context, and this peculiarity characterizes almost all TungusManchu languages with the exception of Northern Tungusic. Thus, in some dialects of Evenki, a new form, in -ni, developed to indicate nominal functions, whereas the old forms became adverbs, i.e. they changed their word class affiliation: boloni “autumn,” bolo “in autumn;” dolboni “night,” dolbo “at night.” Nouns with temporal meaning answer the question atanggi “when?,” “in what time?” which cannot be put to nouns of other semantic groups. Here are some examples of the use of these nouns: cimari ere bou (ma. boo) de n’alma (ma. niyalma) in.the.morning this house DAT people isa-ha gather-PART “In the morning people gathered near this house;” enengi (ma. enenggi) ere ba-de ainu ebu-mbi today this place-DAT why stay-IMPF “Why did (you) stay today in this place?” (SK).
gorel-1.p65
143
4/25/02, 12:23 PM
144
part four
The majority of postpositions, which frequently correlate with case markers, developed from nouns denoting time, place or cause (see Section 8.1). 2.3. Sex of Living Beings In Manchu there is no the grammatical category of gender. The difference between male and female of human beings, animals and birds is expressed by the following means: 1) by the alternation of vowels a and e. The vowel a indicates the male (positive) principle in nature that is yang and occurs in words which denote male human beings, animals and birds. The vowel e indicates the female (negative) principle in nature that is yin, and occurs in words which denote female living beings (see Part 3, Section 6.3.). The following instances show the alternation of the vowels a and e in nouns: ama “father”—eme “mother;” haha “man,” “male”— hehe “woman,” “female;” amaka “husband’s father”—emeke “husband’s mother,” amha “wife’s father”—emhe “wife’s mother;” dehema “hus band of mother’s sister (uncle)”—deheme “mother’s younger sister (aunt);” nakcu “mother’s brother”—nekcu “the wife of one’s mother’s brother;” naca “wife’s elder brother”—nece “the wife of one’s wife’s elder brother;” amila “the male of animals and birds”—emile “the female of animals and birds;” arsalan “lion”—erselen “lioness;” garudai “the male phoenix”—gerudei “the female phoenix;” 2) by combinations of the words haha or hehe and amila or emile with the noun denoting human beings, animals, birds. To indicate the male and female of wild animals the words muhan “a male tiger or panther” and biren “tigress,” “female leopard” are used correspondingly. Also to indicate the female of certain animals the word uniyen “female of certain animals” is used. Here are some instances: haha niyalma “man”—hehe niyalma “woman;” amila temen “a male camel”—emile temen “a female camel;” amila coko “rooster”— emile coko “hen;” muhan tasha “tiger”—biren tasha “tigress;” muhan yarha “a male leopard”—biren yarha “a female leopard;” uniyen honin “a female sheep (ewe);” uniyen ihasi “a female rhinoceros;” 3) by special words denoting male and female individuals: ihan “cattle: bovine, cow, ox, bull”—eje “a castrated bovine,” “ox”—uniyen “a milk cow;” nasin “large black bear”—sati “a male large black
gorel-1.p65
144
4/25/02, 12:23 PM
145
morphology
bear”—nari “a female black bear;” indahån “dog”—enehen/enihen “bitch;” nimaha “fish”—atuha “a male fish”—atu “a female fish;” muà u “quail”—gimà u “a male quail”—bimà u “a female quail.” 2.4. Nouns Denoting Semantics of Quality In Manchu there is a very specific group of nouns denoting various qualitative characteristics of objects and functioning, due to their semantics, as attributes in a sentence (nouns of quality). Due to to the attributive function, they may perform, these nouns are referred to adjectives in most works which study the Tungus-Manchu (and Altaic) languages. It is important to point out that the semantics of quality can manifest itself not only as a qualitative characteristic ascribed to a noun, but also as an abstract concept of certain quality. Such a noun may also refer to a person who possesses a certain qualitative characterictic. Since their main function, which is the attributive one, is not manifested morphologically, the nouns of quality cannot be opposed to other nouns and be defined as adjectives as a part of speech. In accordance with their semantics which combines different meanings, mentioned above, these nouns may carry out different syntactic roles in a sentence. Besides the attributive function, the nouns of quality can modify verbs thus displaying their ability to act as adverbs, mostly as adverbial modifiers of manner. Denoting various abstract notions, they can function as objects as well. These three roles correspond with certain morphological characteristics. In the attributive function the nouns of quality are used in the form of a stem. When functioning as adverbial modifiers, they may be followed by the genitive marker i. When acting as objects, qualitative nouns occur with certain case markers. Due to their syntactic functions they mostly serve as attributes and adverbial modifiers of manner in a sentence, but can be subjects, objects and predicates as well. Occurring in the function of objects, they change their forms in accordance with the requirement of the governing verbs. They are followed by certain case markers, commonly occurring with object nouns, and correlating with the same pronouns ya “what?,” “which?” and ai “what?,” “which?”. Among the nouns of quality several semantic subgroups can be distinguished. The first of them comprises nouns which denote generic notions concerning quality of objects and actions. This
gorel-1.p65
145
4/25/02, 12:23 PM
146
part four
semantic subgroup is represented by the following words: amba “big,” “great,” “large;” “older;” “seniority;” ajige “little,” “small;” “younger;” “the little one;” den “high,” “tall,” “loud;” “height;” ehe “bad,” “wicked;” “evil,” “harm;” “badly;” elehe “peace,” “calm,” “well being;” “peaceful,” “slow;” “easy,” “healthy,” “well;” fe “old,” “not new,” “old aged;” “old times,” “antiquity;” “in old times,” “for merly;” giyan “reason,” “right,” “principle,” “order;” “reasonable,” “right,” “in order,” “proper;” goho “elegant,” “dainty,” “adorned;” “fop,” “dandy;” gosin “love,” “mercy,” “pity;” “charitable,” “gra cious,” “merciful;” labdu “many,” “much;” “numerous;” “a great number;” ice “new,” “fresh;” “the first ten days of the month;” “begin ning,” “the first day of a lunar month;” “at the beginning;” majige “a little,” “a little bit;” “small in numbers;” murin “stubborn;” “stub bornness;” sain “good;” “kind,” “nice,” “pleasant;” “the good,” “well;” àumin “deep;” “depth;” tondo “fair,” “loyal,” “public,” “straight,” “upright;” yargiyan “true,” “real,” “genuine,” “factual;” “truth,” “reality.” One can see that the majority of these nouns denote both qualitative characteristics of nouns and abstract qualitative notions. Correspondingly, they can be translated as substantives and adjec tives. The following examples contain nouns of quality used attributively: niyalma be sain baita yabu-bu-mbi person ACC good affair/matter do/perform-CAUS-IMPF “(They) make persons to do good things” (PASH2:55); tere ajige faha na de caci-ki that small seed earth DAT throw-OPT “Throw that small seed on the earth;” emu amba alim-be dule-he one big mountain(alin)-ACC pass-PART “(They) passed one big mountain” (SK:40). The following examples contain nouns of quality used as subjects and objects: etuhun urse oci ehe be yabu-me powerful people/persons TOP evil ACC make-CONV fafun be neci-mbi law ACC violate-IMPF “As far as the powerful people are concerned (as for powerful people), (they), making evil, violate law” (PASH2:44; MB);
gorel-1.p65
146
4/25/02, 12:23 PM
147
morphology
sain ehe be tua-ha (ma. tuwa-ha) mangi (ma. manggi) good evil ACC see-PART after cimari gene-ki in.the.morning come-OPT “After (I) have seen something good or evil I should come in the morning;” amba ajige hehe haha fiheme isa-fi big small man woman fully gather-CONV “Adults and children (literally: big and small), men and women, all fully gathered together” (SK:40). Nouns of quality may serve as adverbial modifiers of manner: elheken i o-so ume hahila-ra se-he rather.slow GEN become-IMP NEG hurry-PART say-PART “(He) said: “Do (it) slowly, don’t hurry!” (PASH2:97); bi majige muke omi-fi majige amura-ki I some water drink-CONV a.little calm-OPT
“Drinking some water I will calm down a little;”
muke inengdari (ma. inenggidari sain i eye-mbi
water everyday good GEN flow-IMPF
“Water flows well everyday” (SK:40).
Nouns of quality may serve as a predicate:
hafan tere de bolgo oso, baita icih’a-ra (ma. icihiya-ra)
official that DAT honest be(IMP) business do-PART
de tondo oso DAT fair be(IMP)
“If you are an official, be honest; if you are doing business, be fair”
(PASH2:52);
ài (ma. si) encu emu bou (ma. boo) ara-ci sain
you another one house build-CONV good
“It will be good if you build another house” (SK:40).
A subgroup of nouns denoting quality may designate the colour of
objects. These nouns are normally used to indicate attributes or
predicates in a sentence. This subgroup could rightfully be called
adjectives because of their semantics and functions. However, similar
to the majority of nouns having the semantics of quality, they do
not display any morphological characteristics.
To these nouns the following ones can be ascribed: buhe “indigo;”
fulahån “pink,” “reddish;” fulgiyan “red,” “purple;” haksan “golden,”
“reddish brown;” jerde “sorrel (horse);” kara “black (of animals); lamun
“blue;” niohon “greenish” (according to Norman, niohon means
“green”); niowanggiyan “blue,” “green;” sahahån “blakish;” sahaliyan
gorel-1.p65
147
4/25/02, 12:23 PM
148
part four
“black;” sohon “yellowish” (according to Norman, sohon means “deep yellow”); suwayan “yellow;” àahån “whitish;” àanggiyan “white;” àuàu “purple,” “violet;” ulu “white-spotted (horse);” yacin “black,” “dark,” etc. The following examples display the attributive function of the noun of quality in a sentence: emu yacin daimin bi one dark eagle COP “There is one dark eagle;” tere sagal’an (ma. sahaliyan) buga (ma. buha “wild buffalo”) amba that black bull loud jilhan’i suru-me (ma. sure-me) kaica-fi yabu-ha voice-GEN shout-CONV yell-CONV go.away-PART “After bellowing loudly that black bull went away;” yarg’an’i (ma. yargiyan i) meihe muduri labdu, yacin geli reality-GEN snake dragon many dark also bi, àangin (ma. àanggiyan) bi, COP(there.are) white COP(there.are) fulg’an (ma. fulgiyan geli bi, red also COP(there.are) n’ongan (ma. niowanggiyan) sahal’an (ma. sahaliyan green black hacin muduri meihe bi kind dragon snake COP(there.are) “In reality, there are many snakes and dragons; there are dark, and white, and green, and black, there are different kinds of dragons and snakes” (SK:41). Among nouns which denote semantics of quality there are several morphologically marked subgroups. All these nouns are derivative, some of them are formed from object nouns, the others are derived from verbs. They can be regarded as adjectives proper, but some of these subgroups are not numerous, and the suffixes by which they are formed, are not productive. Others include words which denote not only qualitative characteristics of nouns but abstract qualitative notions as well. And what is more important, all morphological markers by which these nominal words are formed, can be found only in the sphere of derivation. These markers do not constitute the adjective as a morphological class of words. Nominal words of one of these subgroups are formed by suffixes
gorel-1.p65
148
4/25/02, 12:23 PM
149
morphology
-hon/-hån/-hun, -shån/-shun. This subgroup comprises words which ascribe certain quality characteristics to nouns. In addition, some words of this subgroup may denote abstract qualitative notions as well: aibishån “swollen,” “a swelling;” banjishån “having sufficient money or goods to lead a comfortable life;” banuhån “lazy;” biyahån “pale,” “wan” (biya “the moon”); bultahån “bulging,” “obvious,” “prominent;” emhun “alone,” “sole,” “lonely;” eneshun “gently sloping;” eshun “raw,” “unripe;” “unripeness;” etuhun “strong,” “powerfull;” “bridegroom;” farhån “dark,” “obscure,” “unclear,” “confused,” “darkness;” fundehun “pallid;” “pallidness;” gehun “bright,” “shining,” “clear;” getuhun “awake;” giltahån “glittering,” “shining,” “clean;” godohon “erect,” “tall and straight;” golmishån “rather long,” “longish;” gosihon/gosihån “bitter,” “miserable,” “suffering;” gulhun “complete,” “intact,” “entire;” hafirahån “narrow,” “pinched,” “tight,” “narrowly;” halhån “hot,” “heat,” “hotly;” jilehun “audacious,” “unabashed;” joboshån “concerned,” “worried,” “concern,” “worry;” juàuhun “sour;” morohon “big and round (of eyes); munahån “morose,” “surly,” “annoyed;” muyahån “entire,” “intact;” oilohon “frivolous,” superficial,” “not serious” (oilo “surface,” “outside”); sitashån “deficient,” “scarce;” “scarcely;” subuhån “sober,” “not drunk;” suilashån “distressed,” “afflicted;” untuhun “empty,” “vacant,” “hollow;” “emptiness,” “space;” wasihån “downward,” “dawn;” “westward,” “to the west” (wasi- “to descend,” “to go down,” “to fall (of rulers),” “to decline (value)”); weihun “alive;” wesihun “up,” “upward,” “east,” “eastward,” “honorable,” “respected,” “superior,” “noble” (wesi- “to ascend,” “to go up,” “to raise,” “to advance (in rank)” ). Zakharov believed that these nominal words (with the suffixes -hon/ -hån/-hun, -shån/-shun) denote a high degree of quality (Zakharov, 1879:83-4). However, it is not observed in regard to all nominal words of this subgroup. These nouns are mostly used in a sentence as attributes: emu alin’i hafirahån kapcu de doài-fi (ma. dosi-fi) one mountain-GEN narrow gorge DAT enter-CONV buce-he bi die-PART COP(there.is)
“Having entered the narrow gorge of a mountain, (she) died.”
bou (ma. boo) i urse dobori gulhun araki omi-fi
house GEN people at.night entire wine drink-CONV
sokto-ho get.drunk-PART
gorel-1.p65
149
4/17/02, 8:57 AM
150
part four
“At night people of the house (servants) drank all the wine and got drunk” (SK:41). These nouns may also serve in a sentence as adverb modifiers of manner: na de oncohon tuhe-he earth DAT on.his.side fall-PART “(He) fell on the earth lying on the back (facing upward)” (SK:41). They may function as predicates as can be seen from the following examples: emu dobori farhån o-ho mangi ... one night dark grow-PART after “One night after it grew dark ... ;” halhån o-fi amba jilgan’i kaica-ha hot become-CONV loud voice-GEN yell-PART “Since it became hot (he) yelled with loud voice” (SK:41). The suffix -ri which may be attached to these nouns, is used to denote both plurality and a high degree of quality: bultahån “bulging”— bultahåri “bulging out (especially the eyes);” fiyangtahån “large and robust,” “a giant,” “a sturdy fellow”—fiyangtahåri “large and robust to a high degree;” gonggohon “deeply thoughtful,” “pensive,” “a deeply thoughtful person”—gonggohori PL of gonggohon; kubsuhun “clumsy,” “large and awkward,” “large and fat,” “a stout/fat man”—kubsuhuri “massive,” “unwieldy,” also PL of kubsuhun; lakdahån “hanging down,” “drooping”—lakdahåri “fully drooping,” “hanging all the way down;” sahahån “blackish,” “rather black”—sahahåri “jet black;” sohon “deep yellow”—sohori “bright yellow.” As one can see, the final -n is deleted before the addition of the suffix -ri. Analyzing this subgroup of nouns one can come to the conclusion that sometimes it is not easy to distinguish substantives from adjectives. Depending on which role they perform in a sentence, these nouns can be referred to as substantives or as adjectives. Correspondingly, the suffix -ri may express the meaning of plurality or a high degree of quality. The suffix -kan/-ken/-kon, used to render the diminutive meaning, may be added to these nouns: eneshun “gently sloping”—eneshuken “somewhat sloping;” eshun “raw”—eshuken “somewhat raw.” Nominal words of quality may be derived from verbs and their forms. They include a subgroup of words ending in the suffix -cuka/
gorel-1.p65
150
4/17/02, 8:57 AM
151
morphology
-cuke. By origin, these nouns are obsolete participial forms which, being constantly used in the role of adjectives, shifted their category status from participles to nouns of quality. These qualitative nouns are referred to as adjectives by some manchurologists. The composite suffix -cuka/-cuke consists of two elements, one of which is the suffix for the participle in -ka/-ke, and the other one arises probably from the old causative suffix -cu which is now obsolete (Lebedeva & Gorelova, 1994:41). This subgroup is not numerous because the suffix by which the words are formed, is not productive: ferguwecuke “wonderful,” “astonishing” (ferguwe- “to be astonish,” “to wonder”); gelecuke “dangerous,” “frightful” (gele- “to fear”); hairacuka “pitiable,” “pitiful” (haira- “to regret,” “to begrudge”); jobocuka “causing concern,” “worrisome,” “distressing” (jobo- “to worry,” “to be distressed”); saiàacuka “praiseworthy,” “worthy of commendation” (saiàa- “to praise,” “to commend”); ulhicuke “understandable” (ulhi- “to understand,” “to comprehend”). Words belonging to this subgroup serve mostly as attributes in a sentence: bi age i ferguwecuke gånin be aina-ha I sir GEN wonderful intention ACC to.do.what?-PART se-me urgede-rakå, urunakå say(AUX)-CONV be.ungrateful.for-PART(NEG) necessarily ujen baili isibu-ki valuable kindness repay-OPT
ainaha seme “surely,” “without fail,” “categorically” (NL:9);
“I shall not surely forget your, sir, wonderful favour, and necessarily
repay your kindness” (PASH2:104);
gelecuke sahal’an (ma. sahaliyan) bou (ma. boo) bi
frightful black house COP(there.is)
“There is a frightful black house” (SK:41).
They also may serves as a predicate:
ere uthai ferguwecuke se-cina this then astonishing say(AUX)-IMP=a note of exclamation at the end of a sentence “This is astonishing!” (PASH2:117); ere daifu bahji-ha durun umuài (ma. umeài) gelecuke this doctor be.born-PART appearance very frightful “This doctor’s appearance (is) very frightful” (SK:42).
gorel-1.p65
151
4/25/02, 12:47 PM
152
part four
Nouns of quality may be formed by the suffix -ba/-be which is added to the verbal stem: kicebe “assiduous,” “diligent” (kice- “to be diligent,” “to exert oneself,” “to concentrate on”); olhoba “careful” (olho- “to fear”); serebe “careful,” “meticulous,” “painstaking” (sere- “to defend,” “to guard (against),” “to prevent”). There are some instances when the suffix -be is added to nominal words: hahiba “quick,” “nimble” (hahi “urgent,” “hurried”). Nouns of quality may be formed by the suffix -su from verbal stems: dahasu “obedient,” “docile” (daha- “to follow,” “to obey”); ejesu “having a good memory” (eje- “to remember,” “to take account of,” “to record”); furgisu “spicy,” “hot;” “ginger” (furgi- “to be hot (of taste)”); gaihasu “suddenly changed,” “unexpectedly better” (gai- “to take,” “to take away,” “to take off”); onggosu “an absent-minded person” (onggo- “to forget”); ulhisu “quick to grasp,” “sensitive,” “keen,” “clever” (ulhi- “to understand,” “to comprehend”). Nominal words of quality may be derived from object nouns by the suffix -ngga/-ngge/-nggo: boconggo “coloured,” “colourful” (boco “colour”); erdemungge “virtuous,” “talented,” “moral” (erdemu “capability,” “virtue,” power”); gosingga “beloved,” “compassionate,” “cherished,” “loving,” “merciful” (dosin “mercy,” “pity,” “love”); horonggo, horonggu “powerful,” “majestic,” “possessing great authority” (horon “power,” “majesty,” “authority”); moringga “pertaining to the horse,” “mounted;” “horseman,” “rider” (morin “horse”). The final -n of a stem is deleted before the addition of the suffix -ngga/-ngge/ -nggo. According to Iv. Zakharov from whose grammar some examples are taken, the suffix -ngga/-ngge/-nggo is the only one by which the proper adjective forms of the nouns, having semantics of quality, are formed in Manchu (Zakharov, 1879:77-8). There is another point of view concerning the nature of this suffix. According to E.P. Lebedeva, originally the suffix -ngga/-ngge/-nggo was used to mark possession. It may be attached to both semantic groups of nouns, viz. to those having the meaning of object in the broad sense and those having semantics of quality: ujungga “first,” “leading” (uju “head,” “first”); ubungge “pertaining parts, portions or shares” (ubu “part,” “portion,” “share,” “responsibility”); hacingga “with distinctions,” “different/various (kinds)” (hacin “kind,” “sort,” “class,” “item”); håsungge “powerful,” “mighty” (håsun “strength,” “power,” “might”); baitangga “business,” “necessary” (baita “matter,” “affair,” “business,” “event”), enduringge “divine,” “holy,” “sacred” (enduri “spirit,” “god,” “deity”), etc. Some of these nouns taking on
gorel-1.p65
152
4/17/02, 8:57 AM
153
morphology
a new qualitative meaning reveal a tendency to shift their category status from object nouns to the class of adjectives: baitangga, boconggo, horonggo, håsungge. In classical Manchu this suffix has three harmonic variants, as opposed to Sibe where, not being subjected to the law of vowel harmony, it has only one variant which is -nge. 2.4.1. The Substantive-Possessive Suffix -ngge Apart from the form of possession in -ngga/-ngge/-nggo, in Manchu there is the substantive-possessive suffix -ngge, which has one harmonic variant. Historically connected to the form of possession in -ngga/ -ngge/-nggo, in the course of time the suffix -ngge became formally and semantically different from the former (Avrorin, 1956:93-100; Lebedeva & Gorelova, 1994:51-2). When added to nouns in the attributive function, the suffix -ngge is preceded by the marker of the genitive i/ni and changes into the formant ningge which is mostly written separately from the noun: ejen ningge “belonging to a ruler, a lord, an emperor.” Even though nominal words, followed by the formants -ngge or -ingge/-ningge, historically had substantive-possessive meaning, in the course of time some of them exhibited a tendency to shift from nouns to adjectives (and be translated as adjectives). It is interesting to point out that the following nominal words are translated as substantives with possessive meaning (for instance, Norman in his Lexicon) or adjectives (Zakharov in his Lexicon and Grammar) by different authors: baingge “of a certain place,” “local;” beyeiningge “one’s own;” dubeingge “the end one,” “the last one,” “last,” “final;” dulimbaingge/dulimbangge “the middle one,” “middle;” feingge “an old thing,” “old,” “ancient;” niyalmaingge “belonging to someone else,” “human;” urseingge “somebody else’s.” When the suffix -ngge follows nouns in the attributive function, it is used to express the superlative degree of quality: tondongge nyalma “the most honest person.” Nominal words in the form of possession in -ngga/-ngge/-nggo may be attached with the substantive-possessive suffix -ngge: amtanggangge “the most tasty”—amtangga “tasty,” “delicious”—amtan “taste,” “smell;” holbonggongge “connected,” “paired,” “married”—holbonngo “connected,” “paired”—holbon “pairing,” “a pair,” “mate,” “mar riage,” “agreement.” Nominal words that end in the suffix -cuka/-cuke also may be followed by the suffix -ngge: ferguwecukengge “the most wonderful,” “the
gorel-1.p65
153
4/17/02, 8:57 AM
154
part four
most astonishing,” “wonder;” ulhicukengge “understanding,” “wise” (Zakharov, 1879:81). Normally the formant ningge follows nominal words, which have the meaning of quality and function as substantives in a sentence, in order to substantivize them. Substantivization is understood here as shifting a nominal word (of any functional and semantic group) or a participle to the grammatical class of substantives. In my opinion, substantivization is different to nominalization. The notion of nominalization is used in a lot of contemporary literature (Aru tyunova, 1976:72; Gak, 1976:85; Paducheva, 1974:193; PDP, 1984:41-6, 174; see also general classifications of the semantics involved in nominalizations made by N.D. Arutyunova, E.V. Paducheva, and E.N. Volf in the book “Semantika i sintaksis,” 1981:202, 102). The term is used to describe the process (and the result of the process) by which a word (and a combination of words) of any grammatical class may function in the noun positions, i.e. as noun analogues. Most nominalizations do not shift their category status to substantives, although some of them do under certain syntactic conditions. In a certain sense, nominalization may or may not be followed by a process of substantivization. In this particular case the whole complex—a noun quality with the formant ningge—denotes a certain object which is the carrier of the corresponding attribute: tacire de amuran ningge “one who is fond of learning;” nure de amuran ningge “one who is fond of wine,” “drunkard” (Zakharov, 1879:80-1). The formant ningge is translated by Norman as “the one which … ,” “he who … ” (NL:213). Here are some examples: gasha bethe foholon ningge dedu-re mangga, bird leg short SBSTR lie-PART capable bethe golmin ningge ele-re mangga leg long SBSTR fly-PART capable Since imperfect participles in combination with the word mangga gain the meaning of usual actions, the following translation of this sentence is correct: “Birds that have short legs usually lie, those that have long legs usually fly” (ORL:181); gemu age-i adali, gucu-se de sain ningge, giyanakå all sir-GEN like friend-PL DAT good SBSTR limited
gorel-1.p65
154
4/17/02, 8:57 AM
155
morphology
udu bi how.many COP
giyanakå udu “how limited?,” “how few?” (NL:110);
“Sir, among friends, are there many good people like you?”
(PASH2:111; QW);
si umesi sain ningge be sabu-hakå o-fi, you(SG) very good SBSTR ACC see-PART(NEG) be(AUX)-CONV teni ere-be sain se-re dabala then this-ACC good say-PART PTL “Since you haven’t seen anything better than this, you consider it as a good thing” (PASH2:113; QW); sain ehe ningge be ilga-bu-rakå good bad SBSTR ACC distinguish-CAUS-PART(NEG) o-ci, bireme gemu sain se-ci geli be(AUX)-CONV completely all good say-CONV again/still o-mbi-o be(AUX)-IMPF-INT se-ci o-mbi-o “is it possible?” “If not to distinguish good (things) from bad ones and call all of them good, is it acceptable?” (PASH2:114). When attached to participles, the substantive-possessive suffix -ngge is used to nominalize the latter (see Section 5.6.5 for a more detailed discussion). 2.5. Diminutives and Augmentatives Diminutives are formed from nominal words by adding the suffixes -kan/-ken/-kon, -gan/-gen, -cen: adalikan “somewhat like,” “rather similar” (adali “like,” “same”); ambakan “rather big,” “a person who is rather large” (amba “big,” “great”); biragan “a small river;” bolgokon “somewhat clean,” “somewhat clear” (bolgo “clean,” “clear”); elheken “rather calm,” “rather well, gentle,” “rather slow” (elhen “peace,” “calm,” “well-being;” “peaceful,” “well,” “slow”); foholokon “rather short” (foholon “short”); gelfiyeken “rather light (of colour)” (gelfiyen “light (of colour);” golmikan “rather long” (golmin “long”); muheliyeken “rather round” (muheliyen “round”); olhokon “rather dry,” “rather thirsty” (olhon “dry,” “dry up”); saikan “pretty,” “good-looking,” “beautiful;” “rather well,” “nicely,” “properly” (sain “good,” “well”); suhecen “a small ax” (suhe “ax”).
gorel-1.p65
155
4/17/02, 8:57 AM
156
part four
When functioning as subjects or objects (direct and indirect), some of these nominal words may take the plural suffix: ambakasi “youths,” “young girls;” ajigesi “small ones.” Diminutives are commonly followed by the genitive case marker i (occasionally they appear in a stem form) and normally function as adverbial modifiers. Due to this, they reveal a tendency to shift from diminutives to adverbs: labdukan i “rather a lot,” “rather many;” saikan i “a little bit better.” According to Zakharov, the word saliyan “just enough,” “just barely enough” being shortened to the suffix -liyan/-liyen when following nominal words, forms diminutives as well: heni saliyan “tiny,” “wee;” adaliliyan “somewhat like,” “somewhat similar;” ambakaliyan “rather large,” “somewhat big” (Zakharov, 1879:82). Following nominal words, the word saka adds the meaning “as if,” “like,” “rather,” “somewhat” to them: ambakan saka “rather large,” “somewhat big;” golmin saka “rather long;” hocikon saka “rather nice;” iletu saka “rather clear;” necikesaka “rather level.” In some cases the word saka is written together with preceding nouns, but sometimes they are written separately. The morphological status of this element is not entirely clear. Obviously, it is the same word that occurring after the imperfect converb is used to denote the meaning “just,” “as soon as” (see Section 5.7.1). According to Zakharov, the word saka is an adverb (Zakharov, 1879:82). In Norman’s view, saka is an adjectival suffix when it occurs with nominal words. But used after imperfect converbs, it should be classified as a clause particle (NL:231). Augmentatives are formed by the suffix -linggå/-linggu: ambalinggå “huge,” “impressive,” “grand;” ehelinggu “very bad,” very wicked,” “very evil.” 2.6. Degrees of Comparison In Manchu the nouns of quality, several subgroups of which are classified as adjectives in other languages, have no morphological category of degrees of comparison. The comparison of two objects regarding their qualitative characteristics, is expressed by a number of syntactic constructions and lexically.
gorel-1.p65
156
4/17/02, 8:57 AM
157
morphology
2.6.1. Comparative constructions The most commonly used construction for the comparative degree includes three semantic and syntactic positions. The first position is occupied by the object to which the other object is being compared. This object is expressed by a noun in the ablative case (ci ). The second position is occupied by the object which is being compared. It is rendered by a noun in the nominative case (zero expression). Finally, the third position is taken by a noun denoting the quality being compared. Here are some examples: manju gisun ci nikan gisun mangga Manchu language ABL Chinese language difficult “The Chinese language is more difficult than Manchu;” i min-ci ahån he I(bi/min-)-ABL older “He is older than me.” In some cases the first position is taken by the object which is being compared: tere ere ci sain that this ABL good “That is better than this.” Sometimes the object which is being compared is not expressed, but it can be reconstructed from the context: min-ci amba
I(bi/min-)-ABL big
“(He) is bigger than me;”
min-ci ahån
I(bi/min-)-ABL older
“(He) is older than me” (PASH:26).
In Manchu there is a comparative construction in which the negative
form of the verb isi- “to reach,” “to arrive,” “to come up to” (isirakå
< isire + akå, which is the negative form of the imperfect participle)
is used. This verbal form requires the dative case from the noun
which denotes the object being compared:
i min-de isi-rakå
he I(bi/min-)DAT approach-PART(NEG)
“He is worse than me” (literally: “He does not approach me”).
gorel-1.p65
157
4/17/02, 8:57 AM
158
part four
Some comparative constructions may include words which denote a high degree of quality, power and superiority: ambula “greatly,” “widely,” “very much;” ele “still more,” “much better,” “especially;” geli “again,” “also,” “still,” “especially;” dabali “exceeding,” “exces sively,” “too;” fulu “excelling,” “surpassing,” “better;” tulgiyen “be sides,” “otherwise.” In such comparative constructions the object to which the other object is compared, is expressed by a noun in the ablative case. Here are some examples: ere niyalma ci geli sain this person ABL still good “(He) is even better than this person;” tere niyalma ci dabali genggiyen that person ABL still.more enlightened “(He) is still more enlightened than that person;” sure genggiyen ulhisu mergen gåwa niyalma ci fulu o-mbi wise enlightened clever skilled other people ABL better be-IMPF “Wise, enlightened, shrewd, and able people are more preferable than others” (PASH:26). The comparative construction may include the nominal word etenggi “hardy,” “powerful,” “strong,” which requires the form of the accusative case of the noun denoting the object which is being compared (Pashkov, 1964:26): naiman i aiman be etenggi etuhun name of people GEN tribe ACC strong powerful “(That tribe) is more powerful than the Naiman tribe;” ere elden àun i elden be etenggi fulu o-hobi this light sun GEN light ACC strong much be-PAST “This light is much stronger that the sun light “(ZAKH:85). The comparison may be rendered by two sentences of opposing meaning. This kind of comparison may be carried out by the construction “anggala ... isirakå.” The postposition anggala “instead of,” “rather than,” “not only” is placed at the end of the first sentence. The negative form of the imperfect participle isirakå taking the last position in the second sentence, requires the form of the dative case of the preceding imperfect participle:
gorel-1.p65
158
4/17/02, 8:57 AM
159
morphology
haksan be yabu-me jabàan be bai-re danger ACC run-CONV advantage ACC look.for-PART anggala, hecin de te-fi hesebun be aliya-ra instead.of calmness DAT live-CONV fate ACC waite-PART de isi-rakå DAT approach-PART(NEG)
“Instead of looking for advantage running the danger (of something),
it is better (literally: “not to approach to...”) to wait (someone’s) fate
living in calmness” (ZAKH:85).
In Manchu there is the comparative construction “ele ... ele” (ele means “still more,” “especially”) which corresponds to the English comparative construction “the more... the more”: ele wesihun, ele cokto still.more respected still.more proud “The more respected, the more proud” (ZAKH:85). The comparison of two objects with regard to a certain qualitative characteristic may be expressed by a special syntactic construction which is similar to the one used in Chinese. This construction includes two sentences, the first of which ascribes a certain quality to the object which is being compared, and the second indicates that the object to which the other object is being compared, does not possess this qualitative characteristic. The following example shows this construction which is considered to be a syntactic calque of the Chinese structure: ere morin sain, tere morin sain akå this horse good that horse good COP.NEG (there.is.not) “This horse is better than that one” (literally: “This horse is good, that one is not good”). In Chinese: Zhe pima hao, na pima bu hao “This horse is good, that one is not good.” 2.6.2. Superlative Constructions The meaning of superlative degree of a noun of quality may be expressed by a syntactic construction which is similar to the com parative one. But although the object to which the other object is being compared is used in singular, the reference is plural, and the noun denoting this object, may be used both in the ablative and genitive cases. Here are some examples:
gorel-1.p65
159
4/17/02, 8:57 AM
160
part four
niyalma ci sain person ABL good “The best of all persons;” niyalma ci wesihun person ABL respected “The most respected person.”
niyalma i sain person GEN good “The best of all persons;” niyalma i wesihun person GEN respected “The most respected person.”
The comparative degree obtains the meaning of the superlative when the object to which the other object is being compared, is expressed by words having the meaning of plurality or a high degree of quality, such as gemu “all,” “in every case;” geren “crowd,” “many,” “many kinds of,” “numerous;” tumen jaka “myriad things;” uheri “general,” “outline,” “summary,” “altogether,” “jointly,” “in general,” “taken as a whole.” Here are some examples: geren ci sain; geren ci amba many ABL good many ABL big “The best one;” “The biggest one;” geren ci ajige many ABL small “The smallest one” (AD:38); tumen jaka ci fulu myriad things ABL better “From all things the best one” (ORL:39). The comparative construction where the nominal word denoting the quality is substantivized by the formant ningge, directly followed by the negation akå, also obtains the meaning of the superlative: ere erin ci oyonggo ninge akå this time ABL important SBSTR COP.NEG (there.is.not) “This time is the most important” (literally: “There is no more important time than this one”); kooli, ere-ci sain ningge akå law this-ABL good SBSTR COP.NEG (there.is.not) gånin, ere-ci jiramin ningge akå thought this-ABL profound SBSTR COP.NEG (there.is.not) “There is nothing better than the law, there is nothing more profound than the thought” (ORL:156).
gorel-1.p65
160
4/17/02, 8:57 AM
161
morphology
The superlative quality may be expressed by repeating nominal words when the first of them is marked by the formant of the genitive case: ujui uju “the first of all;” abkai abka “the most supreme divinity” (literally: “the divinity of all divinities”) (ZAKH:85). The constructions of this type are not numerous. The superlative quality may be formed with the help of words having the meaning of plurality or denoting a high degree of quality, as fulu “excelling,” “surpassing,” “better;” gemu “all,” “in every case;” geren “crowd,” “many,” “many kinds of,” “numerous;” hon “very,” “most,” “too;” jaci “very,” “frequently,” “too;” teni “just,” “truly,” “extremely;” ujui “first,” “primary,” “excessively;” umesi “very,” “to a high degree.” Normally these words precede nouns which denote certain qualitative characteristics: teni amba “extremely big/great;” ujui oshon “excessively cruel;” jaci mentuhun “very stupid;” jaci fahån amba very courage great
“very brave;”
jaci hanci “very close;”
ere niyalma umesi bayan bi
this man very wealthy COP
“This man is very wealthy”(AD:38);
ere niyalma jaci sain akå this man very good COP.NEG (there.is.not) “This man is not good to a high degree” (PASH:27). The idea of superlative quality may be expressed by syntactic constructions including verbs which have semantics of “superiority,” “being outstanding,” “surpassing,” such as the following ones: dabana“to cross over,” “to surpass;” dabatala o- “to act excessively, presumptuously;” dule- “to pass through,” “pass all bounds,” “overstep the limits;” lakca- “to rise,” “to tower above,” “to surpass,” “to be distinguished,” “to be outstanding.” The object to which the other object is being compared, is used in the form of the ablative case: geren ci lakca-ha niyalma crowd ABL surpass-PART man “Prominent (outstanding) man” (literally: “Man who surpassed a crowd”) (Zakh:86). When the object to which the other object is being compared is
gorel-1.p65
161
4/17/02, 8:57 AM
162
part four
singular both in form and meaning, the construction becomes a comparative one: min-ci lakca-ha niyalma I(bi/min-)-ABL surpass-PART man “The man who surpassed me” (AD:38) The verbs colgoro- “excel,” “to surpass” and ete- “to overcome,” “to win,” “to be victorious” also form the constructions having the meaning of the superlative degree. They require the form of the accusative case of a noun which denotes the object to which the other object is being compared: erdemu i geren niyalma be colgoro-ko virtue GEN many man ACC surpass-PART “The most virtuous of all people” (literally: “(Someone) who surpassed all people by his virtues”) (ZAKH:86). 2.7. Declension as a Universal System of Syntactic Subordination. Noun Declension Particular attention should be paid to the category of declension in the Manchu language. The term “declension” has two senses. Firstly, declension is the system of word forms required by the governing verbs. In the second sense, the term “declension” means the capacity of nouns for changing their forms in accordance with the requirement of the governing verbs. Declension has to do with suffixal inflection of word forms. In other words, the markers to indicate cases, should be determined as suffixes. This situation is true of all Tungus-Manchu languages except Manchu. The law of vowel harmony common to all Altaic word forms, does not apply to the case markers in Manchu. Preserving the material proximity to the General-Tungus-Manchu case formants, they reveal their isolation from nouns which they follow. As a result, word forms cannot exist as a synthetic whole. Case markers form analytical constructions with the noun stems in Manchu. The primary function of case forms is to indicate syntactic roles of nouns in a simple sentence. Due to their relative freedom from the word stem, the case markers may occur not only with noun stems, but also with participles. They may indicate not only the relationship of the noun to the verb in a simple sentence, but also express various syntactic relations between principal and subordinate clauses within the complex sentence.
gorel-1.p65
162
4/17/02, 8:57 AM
163
morphology
In Manchu, the category of declension is rather analytical and syntactic than morphological and synthetic. Another point is that Manchu (in comparison with the other languages of the Tungus-Manchu language family) has only a few cases, namely the genitive i, accusative be, dative de, ablative ci and the nominative which has zero expression. Subtleties in meanings cannot be expressed by the existing case markers. To compensate, Manchu has developed analytical combinations composed of case markers and postpositions. Most of the postpositions are derived from nouns with time, place or cause semantics. There is one more case form, though very problematic, which is associated with noun declension, i.e. deri-case form. 2.8. Noun Case Forms (Nominative, Genitive, Accusative, Dative/Locative, Ablative) The marker for the NOMINATIVE (casus indefinitus; following another terminology, the direct) coincides with a noun stem, in other words, this case has zero expression. According to its functions, the nominative (direct) is universal in Manchu. The nominative is one of the principal syntactic cases (same as the accusative and genitive). As it is known, the principal syntactic cases express the semantic roles of agent, patient, experiencer, and stimulus, which are referred to as the participants of situations or events of the outside world. These participants are called arguments which are governed by their predicates (Plungyan, 2000:164, 167). According to E.P. Lebedeva, when the character of the subordination naturally ensues from the semantics of components of a subordinate word-combination it is not necessary to add any case markers to a noun stem in Manchu. The addition of case markers to noun stems becomes obligatory if the character of the semantic relation within the word-combination is unclear (Lebedeva & Gorelova, 1994:44-5). However, it is possible to analyze these language facts in a different way. In Manchu, similar to Turkic languages, the nominative is used with objects which have non-specific (generic) reference. It is assumed by a number of turcologists that morphological categories of case and plural are not obligatory in Turkic languages. Nouns in zero-form, which is normally used to express singular and/ or the nominative, appear in contexts that have semantics of plurality
gorel-1.p65
163
4/17/02, 8:57 AM
164
part four
and require oblique cases. In fact, the rules underlying the use of this zero-form are a more complex. This form is used to express the meaning of “non-specificity” of the grammatical category of definiteness, i.e., it is used with objects which do not have definite or specific reference (Plungyan, 2000:137). A peculiarity of Manchu declension is that government of the case forms of nouns has no automatic or regular character, and its necessity appears only in a certain semantic context. Being normally either the subject of a sentence or the predicate, the nominative is opposed to all oblique cases. But serving to indicate unspecified objects, the nominative may be used parallel to all other case markers. All nouns in the form of the nominative may be used as a subject and a predicate of a sentence: bi hafan o-ki se-mbi I official be-OPT say(AUX)-IMPF “I want to be an official” (ZAKH:127); ere deo umesi hulcin niyalma this younger.brother very foolish person “This younger brother (is) a very foolish person” (PASH1); muke inengdari (ma. inenggidari) sain i eye-mbi water every.day good GEN flow-IMPF “Water every day flows well” (SK). Nouns in the nominative are often used as attributes of all kinds. It was mentioned above that all groups of nouns may serve as attributes. But nouns having semantics of place and quality, as well as numerals normally appear in the form of the nominative when they are used as attributes: si asikan niyalma, bi sagda niyalma You(SG) young person I old person “You (are) a young person, I (am) an old person;” tere bayan niyalma ala-ha that wealthy man tell-PART “That wealthy man told;” duin niyalma casi yabu-me gene-he four person thither go-CONV go-PART “Four persons went thither;”
gorel-1.p65
164
4/17/02, 8:57 AM
165
morphology
fejergi hafasa de afabu-ha under official DAT entrust.to-PART “The under officials have been entrusted (to do something)” (PASH1). When used as attributes, object nouns are usually followed by the genitive marker (i, ni), but may also appear in the nominative case: tere mederi han jili banji-ha that sea khan anger become-PART “That khan of the sea became angry” (SK). Nouns in the nominative often indicate a direct object: giyahån makta-me, indahån be cukule-me falcon launch-CONV dog ACC set.dog.on.prey-CONV amca-bu-mbi chase-CAUS-IMPF “(They) launched falcons and set dogds on prey” (NSB:112); emu hotun (ma. hoton) sabu-mbi one town see-IMPF “(Someone) has seen a town;” emu loho gai-ha one sword take-PART “(Someone) took a sword;” bi àin-de (ma. sin-de) muke bai-me gaji-fi I you(ài/àin-SG)-DAT water seek-CONV bring-CONV omi-bu-ki drink-CAUS-OPT “I shall seek water for you, bring and give to drink” (SK).
Nouns in the nominative function as adverbial modifiers of place: tua (ma. tuwa) doài (ma. dosi) se-ci doài-ki (ma. dosi-ki) fire inside say-CONV enter-OPT muke doài se-ci doài-ki water inside say-CONV enter-OPT “If (you) tell (me) to enter inside the fire (I) shall enter, if (you) tell (me) to enter inside the water (I) shall enter” (SK). Nouns in the nominative may serve as adverbial modifiers of time: soorin de ninju emu aniya te-fi throne DAT sixty one year sit-CONV “sitting on the throne for sixty one years”(ORL:153); ilan aniya sinagala-ha three year mourn-RART
gorel-1.p65
165
4/17/02, 8:57 AM
166
part four
“(Someone) kept the mourning period for three years” (ORL:151); emu inenggi boo aha-si sa-be gama-me one day house slave-PL PL-ACC take-CONV “One day (he) gathered (together) (his) servants (literally: house slaves) …” (NSB:111); cimari bi uka-me yabu-ki morning I run.away-CONV go-OPT “In the morning I shall run away” (SK). The marker for ACCUSATIVE is be, and the law of vowel harmony do not apply to it (see Part III, Section 2). Normally, in classical Manchu it is written separately from a word it follows, with the exception of the irregular pronominal stems (see Section 4.1). But there is evidence that the form gisumbe (< gisun “word,” “language” + be) was also in use in old Manchu (Zakharov, 1879:133). In Sibe the final -n of a noun stem assimilates to the consonant b of the accusative marker, and as a result of the occurrence of morpheme concatenation on the morphemic boundary, a noun stem and the marker be are written in one: sargambe < sargan “woman” + be (see Part III, Section 6). The accusative is also considered to be a principal syntactic case marker. Together with the nominative, it is used to mark participants of situations of the outside world. In Manchu, as in other Tungusic languages (and in all Altaic), the expression of the principal syntactic roles is realized through the so-called accusative (nominative accusative) strategy. That is to say that the case marking of arguments is oriented towards the expression of the opposition between two generalized syntactic roles, namely subject and object. The subject appears in the nominative, and the object is marked by the accusative. In accusative languages, as a rule, the semantic role of experiencer coincides with the role of agent, and the role of patient with that of stimulus (Plungyan, 200:168). The main syntactic function of the accusative is to denote a direct complement of the verb. According to this main role in a sentence, the accusative is used to indicate the direct object of a transitive verb: i boo be weile-mbi he house ACC build-IMPF “He builds a house;” bi hergen be ara-mbi I letter ACC write-IMPF
gorel-1.p65
166
4/17/02, 8:57 AM
167
morphology “I write letters” (ZAKH:159);
emu n’alma (ma. niyalma) bou (ma. boo) be tuak’a-mbi
one man house ACC guard-IMPF
(ma. tuwakiya-mbi)
“A man guards the house;”
ere sagram-be bi gai-ci o-mbi kai this woman-ACC I take-CONV be(AUX)-IMPF MDL.PTL gai-ci o- “(it is) possible to take;” “(I) may take;” “I may take this woman” (SK:48).
In Manchu there is a number of transitive verbs governing the accusative. The corresponding verbs in other languages often may be intransitive. These verbs are: basu- “to laugh at smb., smth.,” “to make fun of,” “to mock at;” daha- “to follow” (the form of the imperfect converb of this verb that is dahame shifted its category status from converbs to postpositions, with the meaning “according to,” “in accordance with”); ejele- “to be master of,” “to occupy by force,” “establish control over;” kadala- “to manage,” “to control,” “to rule;” to kice- “to care about,” “to concentrate on,” “to be intent on;” leole“to talk about,” “to deliberate (on);” se- “to order,” “to command;” sebjele- “to rejoice at smth.;” uile- “to serve,” “to wait on (upon).” The following sentences include such verbs: ai uttu niyalma be basu-mbi what like.this person ACC mock.at-IMPF “How like this (someone) mocks at a person” (ORL:150); fe kooli be dahame yabu-mbi old regulations ACC according.to act-IMPF “(Someone) acts according to old regulations;” ba be ejele-he place ACC occupy.by.force-PART
“(Someone) occupied the place (country) by force;”
sim-be kadala-ra niyalma akå
you(si/sin-SG)-ACC control-PART person COP.NEG (there.is.not)
“There is not a person who controls you” (ZAKH:134);
amba-sa saisa doro- be kice-mbi,
high.official(amban)-PL gentleman way-ACC care.about-IMPF
jemengge be kice-rakå
food ACC care.about-PART(NEG)
gorel-1.p65
167
4/17/02, 8:57 AM
168
part four
ambasa saisa “wise man,” “a true gentleman” (NL:15); “Wise men care about the way, not about the food;” gemu deribun bisi-re be leole-he-ngge all beginning be-PART ACC deliberate.on-PART-NR “deliberation on the being of a common beginning (of things)” (ORL:152); sefu sim-be gene se-he teacher you(si/sin- SG)-ACC go(IMP) order-PART “The teacher ordered you to go” (PASH1:97); sain be sebjele-me, enduri-be leole-me good ACC rejoice.at-CONV spirit-ACC deliberate.on-CONV “to rejoice at good, to deliberate on spirit” (GAB:84); ejen be uile-mbi suzerain ACC serve-IMPF “(Someone) serves the suzerain” (ZAKH:134). In Manchu there are a number of two-placed verbs with mental semantics that govern the accusative. This is unlike other languages having different grammatical systems where certain valencies are filled by nouns in oblique cases: akda- “to depend on,” “to entrust (to),” “to trust;” donji- “to hear about smth.;” gisure- “to speak about smth.,” “to talk about smth;” gåni- “to think about smth., smb.;” hendu- “to speak about smth., smb.” The following sentences include such verbs: mujilen be akda-ra heart ACC entrust-PART
“(I) shall entrust the heart (thoughts, plans, intentions) (to smb.)”
(ORL:149);
tere niyalma be bi akda-mbi that person ACC I trust-IMPF “I trust that person” (ZAKH:133); sin-i algin gebu be donji-fi goida-ha you(si/sin- SG)-GEN fame name ACC hear-CONV endure-PART “It lasted for a long time since (I) had heard about your fame name;” ojo-rakå be gisure-he-ngge do-PART(NEG) ACC speak-PART-NR
ojorakå = it won’t do (NL:222);
“what is being spoken (about) won’t do;”
si mute-rakå bade mim-be ai
you(si/sin- SG) be.able-PART(NEG) if I(bi/min-)-ACC what
gorel-1.p65
168
4/17/02, 8:57 AM
169
morphology hendu-me speak-CONV “If you are not able, what to speak about me?” (ZAKH:134).
The accusative may indicate the material from which something is made: ulgiyan i cecike i funggala be ilha ara-mbi pig GEN small.bird GEN feather ACC flower make-IMPF
ulgiyan i cecike “ kingfisher” (NL:294);
“(They) make flowers with a kingfisher’s feathers;”
ere niyalma na be sangga ara-ha this man earth ACC hole make-PART “This man made the hole from the earth” (ORL:154). The accusative is obligatory when it indicates direct objects governed by transitive verbs in their negative form: cira be tuwa-hakå face ACC see-PART(NEG) “(Someone) did not see the face;” urgun be cira de tuyembu-raku � happiness ACC face DAT reveal-PART(NEG) “(Someone) did not reveal the happiness on a face;” emu gisun be tucibu-me mute-rakå one word ACC bring.out-CONV be.able-PART(NEG) “(Someone) can not be able to bring out not a single word” (ZAKH:134). It may serve to indicate an agent caused to perform the action expressed by the verb in the causative: tere be baita icihiya-bu-me gene that/he ACC business do-CAUS-CONV go(IMP) “Go and order him to do business” (ZAKH:161). The accusative may express space within and means of conveyance by which the motion is going on. It is governed by verbs having semantics of motion: morin be yabu-me horse ACC ride-CONV
gorel-1.p65
169
4/17/02, 8:57 AM
170
part four
“to ride a horse” (PASH1:97); jugån be yabu-me road ACC go-CONV “to go along the road;” bira be doo-me river ACC cross-CONV “to cross a river” (PASH1:104-5). hehe ere alim-be (< alin + be; ma. alin be) tafa-ha woman this mountain-ACC go.up-PART “The woman went up the mountain” (SK:48). The accusative indicates an adverbial modifier of manner: bi suwen-de yargiyan be ala-ra I you (suwe/suwen- PL)-DAT truth ACC say-PART “I say you truly” (literally: “I say the truth to you”); mergen be yabu-re urse ulhi-ci wise/wisdom ACC act-PART people understand-CONV aca-mbi meet/fit(AUX)-IMPF
Tv-ci aca- “should;” “ought,” “must;”
“One should appreciate people who act wisely” (ORL:151).
The accusative may follow participles that serve as predicates in
clauses which are predicative direct complements. This use of the
accusative will be analysed in detail in regards to the predicative
declension of participles (see Section 5.6.3). Several examples follow:
ainu wehe gisure-me mute-re be sabu-hakå why stone speak-CONV may-PART ACC perceive-PART(NEG) ni? INT “Why did you not perceive that stones may speak?” (GAB:112); han’i kåaran’i (ma. kåwarani i) emu ehe sukdun tuci-re khan-GEN stable-GEN one evil spirit appear-PART be sabu-ha ACC perceive-PART
“(They) perceived that an evil spirit appeared in the khan’s stable”
(SK).
It is very important to note that direct objects are not always indicated
by the marker for the accusative.
gorel-1.p65
170
4/17/02, 8:57 AM
171
morphology
According to B. K. Pashkov, the use of the accusative marker depends largely on the position of the direct object in a sentence. When a direct object is placed directly before a governing verb, the marker for the accusative can be left out. Any part of a sentence, i.e., subject, indirect object or adverbial modifier of manner may be placed between the direct object and its governing verb. To describe such a position of the direct object, Pashkov introduced into practise the term “distant.” It means that the direct object appears rather far removed in the sentence from the verb that governs it. In such syntactic context the marker for the accusative is obligatory: tere niyalma be bi akda-mbi that man ACC I trust-IMPF “I trust that man ” (ZAKH:133); we sim-be gene se-he who you(si/sin- SG)-ACC go(IMP) say-PART “Who ordered you to go?” (ZAKH:134); hergen be saikan ara letter ACC good write(IMP) “Write letters better” (PASH1:98-9). It seems that Bosson’s description of the use of the accusative in contemporary Mongolian is close to that in classical Manchu as described by Pashkov. Bosson believes that when an inanimate direct object is close to the verb and is clearly the object of the action, there is no necessity in using the accusative, and the indefinite (nominative, direct—L.G.) is often used instead (Bosson, 1964:28). According to A. Kim, in contemporary Mongolian the accusative is used with the definite object NPs, and the absence of the accusative marker signals that the NPs are used to convey indefinite objects (Kim, 1988:164). Analyzing the Mongolic Buryat grammatical tradition, E.K. Skribnik suggests that the opposition of marked and unmarked NPs, which are used to express direct objects, can be explained in terms of definiteness vs. indefiniteness. In her opinion, this opposition is also closely connected with discourse structuring in Buryat, in which the accusative is used to mark a secondary clausal topic (Skribnik, 2001). As mentioned above, in Turkic languages, the accusative is used
gorel-1.p65
171
4/17/02, 8:57 AM
172
part four
with direct objects which have specific reference (definite objects). In the Turkish language, for example, the opposition of marked and unmarked NPs, which are used to express direct objects, reflects the semantic opposition between definite vs. indefinite or specific vs. non specific direct objects ((Nilsson, 1978-1979; Johanson, 1998 respectively). The opposition between specific vs. non-specific direct objects, as it has been suggested by L. Johanson, is closely connected with the means of pragmatic structuring in Turkish. Analyzing grammatical traditions of some Turkic languages of Siberia, namely Turkic Yakut and Altai Turkic Telengit, Skribnik suggests that the opposition between marked and unmarked NPs in the direct object position can be explained on the basis of definiteness vs. indefiniteness. In the Yakut language, in which five cases may occur in the direct object position, the accusative suffix is used to mark a secondary clausal topic, whereas the nominative (casus indefinitus) is used to mark rhematic unfocussed objects, and the partitive usually marks rhematic focussed objects (Skribnik, 2001; for more detailed information concerning pragmatic structuring in languages, including Manchu, see Part V, Section 3). According to Norman, in the important Manchu dialect, Sibe, the accusative is used with direct objects which have definite or specific reference (Norman, 1974:166). As far as classical Manchu is concerned, there is no strict explanation as to when the marker of the accusative is obligatory and when it can be omitted, and the problem needs to be fully investigated. Although it is clear that the appearance of the accusative is closely related with the category of definiteness. Several patterns can be distinguished where the accusative is not used in Manchu: 1) a direct object and a governing verb are formed from the same root and because of that have the same semantics: etuku etu- “to put on clothing” (etuku “clothing;” etu- “to put on [clothing]”); gucu gucule“to make friends,” “to be friends with” (gucu “friend;” gucule- “to make friends”); irgebun irgebu- “to sing songs,” “to compose verse” (irgebun “song;” “verse;” irgebu- “to sing songs,” “to compose verse”) (GAB:136); ficakå fica- “to blow a flute” ( ficakå “flute;” fica- “to blow [a flute]”); makcin makci- “to dance” (makcin “dance;” makci- “to dance”); nirugan niru- “to paint a picture” (nirugan “picture,” “painting;” niru- “to paint,” “to draw”) (ORL:154).
gorel-1.p65
172
4/17/02, 8:57 AM
173
morphology
However, there is one exception to this rule as shown by the following example: jalan i sebjen be sebjele world GEN joy/happiness ACC to.rejoce “to rejoce at happiness of the world” (GAB:136). 2) in a string of direct objects only the last of them is marked by the accusative: eme ama be mafa se-mbi mother father ACC ancestor call-IMPF “A mother and a father are called ancestors (mafa)” (ORL:150); 3) direct objects denoting means and ways of conveyance and governed by transitive verbs: jugån nei-me road/way open-CONV
“to open a road (way)” (PASH1:105); “to open a way (by force)”
(NL:164).
4) in subordinate clauses predicates expressed by participles may
govern direct objects not indicated by the marker of the accusative:
baita icihiya-ra de tondo oco business do-PART DAT honest be(IMP) “When doing business, be honest” (PASH1:111). In Manchu there is a number of two-placed verbs the first valency of which is replaced by the direct object not indicated by the accusative. These verbs are: ara- “to do/make smth. with smth.;” weile- “to make smth. with smth.;” obu- “to consider smb./smth. as smb./smth.;” se- “to call smb./smth. by smb./smth.” Here are some examples with the above verbs: be sim-be moo-i uren se-mbi? who you(si/sin- SG)-ACC wood-GEN statue call-IMPF “Who call you by wooden statue?;”
gorel-1.p65
173
4/17/02, 8:57 AM
174
part four
ende-bu-he be hala-rakå o-ci err-PASS-PART ACC correct-PART(NEG) be(AUX)-CONV tere be endebuku se-mbi that ACC mistake call-IMPF “If do not correct errors this is called a mistake” (ORL:150); ai be fulehe da obu-mbi ai be temgetu what ACC root base consider-IMPF what ACC proof
obu-mbi?
consider-IMPF
“What is considered as a base, what is considered as a proof?”
(ZAKH:133);
bireme gemu tere be sain se-mbi totally all that/he ACC good call-IMPF “Totally all consider him as a good (person)” (PASH1:101); na be sangga ara-fi tomo-mbi earth ACC hole make-CONV live-IMPF “Made a hole from the earth, (they) live (in it)” (ORL:154). When the verb se- has the meaning “to say,” “to speak,” and the whole construction is used in the meaning of the Latin “accusativus cum infinito,” the marker of the accusative indicates the only position which is replaced by a noun having the meaning of a person or by a personal pronoun: mim-be gisure-re se-mbi I(bi/min-)-ACC speak-PART say-IMPF “It is said (they say) that it was me who spoke;” enen sira-bu-re jalin be progeny continue-PASS/CAUS-PART on.account.of ACC niyalma-i tuhen se-mbi person-GEN purpose say-IMPF “It is said (they say) that continuation of progeny is any person’s purpose” (ORL:154). Finally, there can be several direct objects indicated by the accusative marker in a single sentence in Manchu. It is could be explained by the fact that verbal nouns preserve verbal government and themselves may be governed by verbs. The following examples display the occurrance of several markers of the accusative in a single sentence:
gorel-1.p65
174
4/17/02, 8:57 AM
175
morphology
waliya-ha usin be suksala-bu-re be abandon-PART field ACC open.to.cultivation-CAUS-PART ACC hese-i yabubu-re jalin order-GEN carry.out-PART in.order.to “in order to carry out the order to open to cultivation of abandoned fields” (ZAKHL: 967). There are two direct complements in this sentence, the first suksalabure be “opening to cultivation” and the second usin be “field for cultivation.” se-re-ngge habàa-ra be akåbu-ki se-ci say-PART-NR accuse-PART ACC stop-OPT say-CONV tere-i that-GEN = their deribun be lashala-ra be wesihun beginning ACC terminate-PART ACC honourable obu-re be hendu-hebi consider.as-PART ACC say-PAST Tv-ki se- “to wish to do smth.,” “to intend to do smth.;” “It is said that he who wishes to stop accusations should consider important to prevent their beginning” (ORL:154). The principle predicate hendumbi “(they) say” governs the predicative direct object expressed by the participle in the accusative which is obure be “that (one) should consider.” The latter, in its turn, governs the predicative direct object which is expressed by the participle in the accusative lashalara be “to terminate,” “termination” which governs the noun in the accusative deribun be “beginning.” The marker for the GENITIVE is i, ni. When nouns are changing in accordance with the requirement of the verb, the final -n of the stem is usually deleted. But in the genitive the final -n is preserved. The marker i is used after words ending in a vowel or in -n. Words ending in a consonant other than n, are followed by ni: gurung ni doro “inside the palace” (ZAKH:128). After words anding in i, especially those borrowed from Chinese, the marker for the genitive may be omitted (Möllendorff, 1892:5). There is evidence that in Sibe, if the noun stem has final -ii, then the genitive marker is zero (Norman, 1974:5). In most cases the marker for the genitive is written separately from words it follows: alin i, gisun i, boo i, gurun i, juse i, mujilen i, na i. Occasionally words and the marker for the genitive i following them are written as one: abkai ( abka “sky,” “heaven” + i), biyai (biya “moon,”
gorel-1.p65
175
4/17/02, 8:57 AM
176
part four
“month” + i), booi (boo “house” + i), erei (ere “this” + i), gisuni (gisun “speech,” “word,” “language” + i), hådai (håda “business,” “good” + i; hådai ba “market”), hiyalmai (niyalma “man,” “person” + i). The genitive is also considered to be one of the principal syntactic cases. Its primary syntactic function is to express an argument which is governed by a noun. Expressing only one, but a very general syntactic relationship between two nouns, the genitive case accu mulated a lot of various semantic functions in the Indo-European languages (Benveniste, 1961). It should be mentioned that in Turkic languages the use of a subordinate noun without the genitive marker indicates a non-specific interpretation (Plungyan, 2000:186). The primary semantic function of the genitive is the possessive one, according to which it indicates the possessor of an object, its origin and habitation: boo i ejen “the master of the house;”
house GEN master
abka-i ejen “the Lord of Heaven, God” (MOLL:5);
heaven-GEN lord
abka-i jui “the Son of Heaven, the Emperor”
heaven-GEN son
biya i elden “the light of the moon;”
the moon GEN light
in-i jaka “his thing;”
he(i/in-)-GEN thing
men-i baita “my business;”
I(bi/min-)-GEN business
gurun i dasan “the government of the state;”
state GEN government
jui i banin “the character of the son” (ZAKH:128).
son GEN character
The genitive may also indicate:
1) part of a whole:
alin i ninggu “the top of a mountain” (ZAKH:128);
mountain GEN top
buhå i weihe “horns of a Manchurian deer;”
deer GEN horn
gorel-1.p65
176
4/17/02, 8:57 AM
177
morphology temen i uju “the head of a camel;”
camel GEN head
ihan i àan “the ears of a cow;”
cow GEN ear
meihe i meifen “the neck of a snake” (ORL:136-7);
snake GEN neck
muke i sekiyen “river-head” (SK:47);
river GEN source
2) material from which a thing is made:
yacin dobihi i dahå black fox.pelt GEN fur.coat “a fur coat [made] from pelt of black fox;” fulgiyan erimbu wehe i jingse red precious stone GEM knob.indicating.rank.on.an.official.cap “a knob [made] from red precious stone” (ORL:141); wehe i kåaran (ma. kåwaran) “a stable made from stone (a stone stone GEN workshop stable)” (SK:47); 3) quantity, measure, and time: emu moro i àahårun muke omi-ha one cup GEN cold water drink-PART “(Someone) drank one cup of cold water;” nadaci jalan i omolo the seventh generation GEN grandson “Grandson seven times removed;” ududu biya-i caliyan several month-GEN pay.and.provisions “pay and provisions for several months” (ORL:138); 4) persons according to their kinship and subordination: eme i jui “the mother’s son” (NSB:114); mother GEN son han i sargan “the khan’s wife;” khan GEN wife
gorel-1.p65
177
4/17/02, 8:57 AM
178
part four
ama-i jui “the father’s son (child)” (SK:47);
father-GEN son
5) intention with which a thing (or event) is done:
urgun i doro “the ceremony of congratulation”
congratulation GEN ceremony
(MOLL:5).
Possessive relations may link an agent and the action this agent
performs. Possessive relations may also link a patient and the current
state of the patient (or the change the patient undergoes) as well as
an experiencer and the action of perception of visual or auditory
information. In this case, the marker of the genitive follows a word
which denotes a subject of clauses of certain kinds:
na i tuci-ke jaka earth GEN originate.from-PART thing “things which originated from the earth” (ZAKH:128); fudzi i hendu-he donji-ha-ngge Confucious GEN say-PART heard-PART-NR “(They) heard what Confucious said” (ORL:140); dziwen uttu oci terni yargiyan mujilen i proper noun if so his honest heart GEN gurun i jalin faààa-ha-ngge kai country GEN for.the.sake.of exert.effort-PART-NR COP “As far as Dziwen is concerned, his honest heart cares for the sake of fatherland” (ORL:135); in-i takåra-ha niyalma he(i/in-)-GEN delegate-PART person “the man delegated by him” (ZAKH:129). The attributive function, in its broadest sense, has developed from the possessive function, which is the primary semantic meaning of the genitive. Nouns followed by the marker of the genitive normally serve as attributives in a sentence: doro-i umiyesun “a belt for a court dress or ceremonial garment;” rite-GEN belt doro-i yoro “a ceremonial arrow” (NL:63); rite-GEN arrow
gorel-1.p65
178
4/17/02, 8:57 AM
179
morphology
dorolon i nomun “the book of rites” (PASH1:76); rite GEN classic.book urgun i cira “a merry/lively face” (ZAKH:128); joy GEN face tuba i niyalma ji-he that.place GEN people come-PART “The people of that place came” (ZAKH:309); fe ice i baita, uru waka i arbun old new GEN bussines right wrong GEN circumstances dursun form/pattern/shape
“old and new businesses; right and wrong circumstances”
(PASH2:57).
Occurring after nouns with semantics of objects, the genitive is used
in combinations of word which become fixed nominal units:
bithe-i niyalma “scholar;”
book-GEN person
cooha-i niyalma “soldier,” “warrior;”
soldier-GEN person
golo-i amban “provincial officials of high rank;”
province-GEN high.official
monggo i ba “Mongolia” (PASH1:89);
Mongolia GEN place
håda-i niyalma “merchant” (ZAKH:128);
business-GEN person
håda i ba “market,” “marketplace;”
business GEN place
håda i jaka “merchandise” (NL:139);
business GEN thing
abka-i fejergi “all under heaven,” “the world,” “China”
heaven-GEN underneath
(ORL:136).
All of the above are listed as stable word-combinations in Norman’s
Concise Manchu-English Lexicon.
In Manchu, one sentence may include a sequence of attributes in one noun phrase when attributes ascribe certain qualitative characteristics to the preceding attributes:
gorel-1.p65
179
4/25/02, 12:28 PM
180
part four
julge-i da-i wang sa-i mujilen i ancient.times-GEN leader-GEN monarch PL-GEN intention GEN doron be sa-ci o-mbi way.of.thinking ACC know-CONV become-IMPF sa-ci o- “(it is) possible to know;” “It is possible to know the way of thinking of monarchs of ancient times” (PASH1:92). The genitive is used to attribute qualitative chatacterictics not only to nouns, but to participles and verbs as well. Occurring after nouns having qualitative semantics, the genitive indicates the adverbial modifier of manner: giyan fiyan i gama-mbi funcen daban i reason appearance GEN arrange-IMPF extra GEN icihiya-mbi put.in.order-IMPF
giyan fiyan “reasonable,” “orderly,” giyan fiyan i “in proper order,”
“in detail” (NL:110);
funcen daban “extra and excess,” funcen daban i “excessively” (NL:95);
“(Someone) has arranged (deels) properly, and put (everything) in
order excessively” (PASH2:58);
tere sahal’an (ma. sahaliyan) buhu (ma. buha) amba jilgan’i
that black bull loud voice-GEN
sure-me (ma. sure-me) kaica-fi yabu-ha
shout-CONV yell-CONV go-PART
“That bull yelled with loud voice (loudly) and went away;”
ama eme be saikan’i gingule-me (ma. ginggule-me)
father mother ACC rather.well-GEN respect-CONV
banji live(IMP)
“Respecting your father and mother, (let you) live well” (SK:47);
bi sain mujilen i niyalma be tuwa-ra o-ci, I good intention GEN people ACC look.at-PART be(AUX)-CONV niyalma urunakå sain mujilen i mimbe tuwa-mbi people surely good intention GEN I(bi/min-)-ACC look.at-IMPF “If I look at people with good intention then people surely look at me with good intention” (ZAKH:129). Adverbs are normally formed by repeating the noun followed by the genitive marker: cun cun i “gradually;” dahån dahån i “repeatedly”
gorel-1.p65
180
4/17/02, 8:57 AM
181
morphology
(dahån “again”); giyan giyan i “in proper order” (giyan “order,” “proper”); siran siran i “continually” (siran “continuation”); ulhiyen ulhiyen i “gradually” (ulhiyen “gradual”) (ZAKHL, NL). The genitive may also denote the following additional meanings: 1) distributive: geren saisa i uju o-me mute-hebi all scholar GEN the.first become-CONV be.able-PAST “He was able to become the first of all scholars” (ORL:137); gucu gucule-re-ngge niyalma-i sunja ciktan i friend make.friends-PART-NR people-GEN five principle GEN emke one (as substantive)
“The fact that people make friends is one of the five principles of
people” (ZAKH:128-9);
2) instrumental:
angga i fodoro-me mouth GEN pant-CONV “to pant with mouth;” bulukan i muke i silgiya-ra warm GEN water GEN rinse-PART “rinsing with warm water” (ORL:140); ere suhe-i saci-mbi this axe-GEN chop-IMPF
“(I) shall chop with this axe;”
tere futa-i anga (ma. angga) be håaita-fi (ma. håwaita-fi)
that rope-GEN hole ACC tie.up-CONV
“tied up a hole with that rope” (SK:47);
amba-sa saisa niyalma i
high.official(amban)-PL scholar/gentleman people GEN
niyalma be dasa-mbi
people ACC rule-IMPF
ambasa saisa “a worthy, wise man,” “a true gentleman” (NL:15);
“Wise men rule over people by means of people” (GAB:65).
In Manchu, nouns are normally followed by the genitive before a
number of postpositions. The genitive marker does not in this case
gorel-1.p65
181
4/17/02, 8:57 AM
182
part four
express a certain kind of a relation between the noun and the
postposition, it merely serves is a formal link between them. Here
are some examples where the genitive marker serves as a formal
device linking a noun and a postposition following it:
bira-i cala “on the other side of the river;”
river-GEN on.the.other.side
mujilen i dolo “in a heart;”
heart GEN inside
sin-i emgi gisure-ki
you(si/sin- SG)-GEN with speak-OPT
“(I) shall speak to you;”
sin-i baru ala-ki
you(si/sin- SG)-GEN to tell-OPT
“(I) shall tell you;”
tere-i jalin gisure-he this/he-GEN for.the.sake.of speak-PART “(Someone) spoke for the sake of him” (ZAKH:128-130); yao sun i forgon i gese proper name proper name GEN time GEN like “like at the times of Yao and Sun” (ORL:138). The marker for the DATIVE is -de. Like the accusative, the dative does not follow the law of vowel harmony. In most cases the marker of the dative is written separately from the word it follows, with several exceptions to the rule. The marker and irregular forms of personal pronouns are normally written as one. The marker de and the copula bi are also written as one in the following patterns: amargi debi “it is situated in the north;” dergi debi “it is situated in the east;” julergi debi “it is situated in the south;” wargi debi “it is situated in the west.” Like the accusative, the dative plays a very important role in the structure of Manchu utterances. Although the dative is oriented towards the expression of the peripheral semantic roles of arguments, it is used to express many of these. The dative is primarily used to indicate the semantic role of recipient, which is normally combined together with the role of addressee and in some languages with the semantic roles of benefactive and experiencer. All of these semantic roles are generalized by the case for the syntactic indirect object, i.e. the dative (Plungyan, 2000:169-70). The semantic role of
gorel-1.p65
182
4/17/02, 8:57 AM
183
morphology
instrument, which is considered to be a peripheral argument, is also marked by the dative in Manchu. Being a polyfunctional case, the dative is used in Manchu to indicate circumstantial semantic roles, primarily, to express space and temporal relationships. Corres pondingly, the dative has two basic syntactic functions: it is used to indicate an indirect object and adverbial modifier, mostly of place and, less frequently, time and cause. Indicating an indirect object, the dative refers to the recipient of the action or the addressee of speech or another action: ere niyalma de bu-mbi this man DAT give-IMPF “(Someone) gives to this man;” tere niyalma de hendu-he that man DAT say-PART “(Someone) said to that man” (MOLL:5); abka-de hengkile-mbi heaven-DAT kowtow-IMPF “(I) kowtow to heaven” (ZAKH:130); ere sunja n’alma (ma. niyalma) tere aàikta (ma. asihata) de this five man that guy DAT fonji-ha ask-PART
“These five men asked that guy;”
wan be dang de jaka hacin hacin boy bi (ma. boobai)
proper name DAT thing various various treasure
ulin ulha adun gemu labdu bu-fi property cattle cattle all many give-CONV wan be dang = proper name; “Wan Be Dang was given all kinds of things—treasure, property and cattle” (SK:48). The dative may indicate the agent of a passive verb: bi in-de gele-bu-he I he-DAT fear-PASS-PART “I was scared of him” (I was frightened by him) (ZAKH:162); … julergi gurun joo de wa-bu-fi senior ruling.house zhao DAT kill-PASS-CONV “(They) were killed by the Senior Zhao ruling house;”
gorel-1.p65
183
4/17/02, 8:57 AM
184
part four
amargi joo gurun de mukiye-bu-hebi junior zhao ruling.house DAT extinguish-PASS-PAST “(It) was extinguished by the Junior Zhao ruling house” (ORL:145). The dative may indicate a person who is the possessor of something. A predicate of such a sentence is expressed by the verb bi- “to be,” “to exist” (or by the copula bi), which takes the meaning “to have,” “to possess” in such constructions as following: ahun de bithe bi elder.brother DAT book COP(have/there.is) “The elder brother has a book;” bithe-i niyalma usin i haha weile-re faksi book-GEN person field GEN man work-PART workman
håda-i urse de men-i
business-GEN people DAT we(be/men-EXCL)-GEN
men-i ula-ha tacin bi
we-GEN pass.on-PART skill COP
bithei nyalma “scholar,” usin i haha “tenant farmer,” weilere faksi
“workman,” hådai niyalma “merchant” (NL:31, 300, 138);
“Be he scholar, farmer, workman or merchant, every man has some
skills to pass on” (ORL:147);
min-de gisun bi-fi sin-de fonji-ki
I(bi/min-)-DAT word be-CONV you(si/sin- SG)-DAT ask-OPT
se-mbi
say(AUX)-IMPF
Tv-ki se- “to want/wish to do smth.;”
“I have a word, (I) want to ask you about” (literally); “There is
something I want to ask you about”(ZAKH:131).
The dative may indicate parenthetical words denoting the source
of a statement (dictum) or quotation:
min-i gånin de “in my opinion” (ZAKH:144);
I(bi/min-)-GEN thought DAT
ejen i hese de “according to the emperor’s edict”
emperor GEN edict DAT
(ZAKH:131).
The dative also indicates an instrument of the action:
gorel-1.p65
184
4/17/02, 8:57 AM
185
morphology angga de håla, mujilen de eje mouth DAT read(IMP) mind DAT memorize(IMP) “Use your lips to read and your brain to remember;” tere-i elden de eldeàe-me that-GEN light DAT shine-CONV “Being illuminated by that light … ” (ORL:145).
Both the genetive and the dative are used to indicate an instrument, but there is a difference between them. The dative is used with verbs in the past: beye-i gala-de jafa-habi self-GEN hand-DAT catch-PAST “(Someone) caught with his hand;” beye-i gala-i gaisu self-GEN hand-GEN take(IMP) “Take with your hand” (ZAKH:132). The dative is used when people talk about someone else’s activities: si yasa-de twva-ki you(si/sin- SG) eye-DAT see-OPT “You see with your eyes”(ZAKH:132). But when the question is about someone’s own activity, then the genitive is used: min-i yasa-i tuwa-ci ... I(bi/min-)-GEN = my eye-GEN see-CONV “Seeing with my own eyes ... ” (ZAKH:132). Serving as an adverbial modifier of place, time, and cause the dative may show: 1) location in space (in, at): alin bujan-de tomo-mbi mountain forest-DAT live-IMPF “(They) live in mountains and forests;” kuwecihe boo-de uji-mbi pigeon house-DAT nurture-IMPF
gorel-1.p65
185
4/25/02, 12:24 PM
186
part four
“A pigeon is nurtured by a household” (ORL:146); abka de deye-re gasha bi sky DAT fly-PART large.bird COP na de feksi-re gurgu bi earth DAT run-PART wild.animal COP “Those flying in the sky are birds: running on the earth are wild animals” (ZAKH:131); bi ere ba de àim-be (ma. sim-be) ala-me I this place DAT you(ài/àin- SG)-ACC wait.for-CONV ningun (ma. ninggun) an’a (ma. aniya) o-ho six year become-PART “Six years passed since I began waiting for you here (in this place)” (SK:47); 2) destination (towards, upon, on): tugi de sucuna-me deye-mbi cloud DAT soar-CONV fly-IMPF “(They) fly soaring towards the clouds” (ORL:147); ce boo-de isinji-ha they house-DAT reach-PART “They reached the house;” mukden de gene-mbi place name DAT go-IMPF
“(I) go to Mukden” (ZAKH:131);
tere bou (ma. boo) i uce de yabu-me iàina-ha (ma. isina-ha)
that house GEN door DAT do-CONV arrive.at-PART
“(Someone) arrived at the door of that house” (SK:47);
3) location in time (it should be mentioned that in a simple sentence
an adverbial modifier of time is more often expressed by the
nominative):
tere nergin-de “at that time,” “at that moment” (ZAKH:131);
that time/moment-DAT
uju-i biya-de “at the first moon” (ORL:147);
first-GEN moon-DAT
muda-ra erin-de ere ba-be boljon obu-ki return-PART time-DAT this place-ACC agreement make-OPT “Let us make an agreement when to return to this place” (SK:48);
gorel-1.p65
186
4/17/02, 8:57 AM
187
morphology
4) the limits of a certain length of road or space of time (from... to, from ... until): golo-i hecen ci gemun hecen de province-GEN city ABL capital city DAT “from a provincial city to the capital” (ZAKH:131); julge-ci te de isitala ancient.times-ABL now DAT until “from ancient times until the present” (ORL:144); 5) as far as the temporal meaning of the dative is concerned, it is mostly found in complex sentences, in clauses expressing semantics of time. Following participles, the dative indicates a general temporal meaning: amba-sa teike gene-re de ala-ha high.official-PL the.moment go-PART DAT report-PART “The moment the high officials were about to go, (he) reported to them;” tuwa-ra de ja gojime, yabu-re de mangga look.at-PART DAT easy though perform-PART DAT difficult “Though easy to watch, it is difficult to perform” (ZAKH:131); in-i kemuni leole-me gisure-re-de ... he(i/in-)-GEN still discuss-CONV speak-PART-DAT “when he was still discussing ... ” (ZAKH:132); ama-i bisi-re-de tere-i gånin be father-GEN be/exist-PART-DAT that-GEN opinion ACC tuwa-mbi look-IMPF “When father was alive, (they) listened to his opinion” (PASH2:106); 6) the dative may indicate an adverbial modifier of manner: jobolon de “sorrowfully,” urgun de “joyfully,” doron de “according to rites,” “ceremonially;” 7) cause: emu gisun de mergen obu-mbi, emu gisun de mergen one word DAT wise make-IMPF one word DAT wise
gorel-1.p65
187
4/17/02, 8:57 AM
188
part four
akå COP.NEG (there.is.not)
“One word can give rise to wisdom or foolishness (ORL:144).
The dative is usually governed by nouns and verbs having semantics
of evaluation such as ja “easy,” mangga “difficult,” sain “well,” amuran
“fond of,” “intent on,” duibule- “to compare,” etc.:
sa-ra de ja learn-PART DAT easy “it is easy to learn;” sure-be yabu-re de amuran wisdom-ACC perform-PART DAT intent.on “being intent on performing wisely” (ORL:143); arsari urse de duibule-ci ojo-rakå ordinary people DAT compare-CONV be(AUX)-PART(NEG) “(He) cannot be compared with ordinary people” (ORL:144); yabu-re de mangga perform-PART DAT difficult “it is difficult to perform;” jete-re de amuran, yabu-re de banuhån eat(je-)-PART DAT good.at work-PART DAT lazy “Being good at eating, but lazy to work” (literally); “Being greedy for food but with no appetite for work” (ZAKH:131). The verbs that have emotional semantics such as akda- “to trust;” daya- “to be dependent on,” “to rely on;” gele- “to fear;” giru- “to feel ashamed;” golo- “to be startled,” “to be scared;” gucule- “to make friend;” jaila- “to avoid,” “to get out of the way of;” jobo- “to suffer;” sengguwe- “to fear,” “to dread;” ubiya- “to loathe;” targa- “to abstain from” also govern the dative: muten akå de jobo-mbi capability there.is.not DAT suffer-IMPF
“(Someone) suffers when there is no capability (to do smth.)” (literally);
“Incapability brings suffering” (ORL:143);
faksi anggalinggå de gucule-re de o-ci shrewd glib DAT make.friends-PART DAT be(AUX)-CONV ekiyende-mbi be.lacking-IMPF
gorel-1.p65
188
4/17/02, 8:57 AM
189
morphology
“When making friends with glib (persons) then (someone) will be deficient” (literally); “Glib friends are unsatisfactory;” boco becun de targa lust quarrel DAT avoid(IMP) “Lust and quarrels are best avoided;” fusihån de fonji-re de giru-rakå junior DAT ask-PART DAT be.ashamed-PART(NEG) “There is no shame in (asking) taking advice from the young” (ORL:144); abka de ubiya-fi heaven DAT loathe-CONV “being loathed by Heaven” (ZAKH:130); in-i donji-rakå de gele-me golo-mbi he(i/in-)-GEN hear-PART(NEG) DAT fear-CONV be.scared-IMPF “He is afraid of not being able to hear” (PASH2:41; MB). The dative is also governed by the following verbs: aca- “to meet,” “to meet together,” “to correspond to,” “to fit,” “to be equal;” aliya“to wait;” bai- “to ask for;” donji- “to listen,” “to hear;” fonji- “to ask;” holbo- “to get married;” isina- “to reach;” taci- “to learn,” “to study,” “to become accustomed to.” Here are some examples with these verbs: in-i adaki de bai-fi bu-habi he(i/in-)-GEN neighbour DAT ask-CONV give-PAST “(He) returned what he had borrowed from his neighbour” or “Having borrowed from his neighbour (he) made repayment” (ORL:146); duin erin-de acabu-me four season-DAT adapt-CONV “adapting to four seasons;” erin de aliya-hai time DAT wait-CONV “having waited for the time” (ORL:144); sin-de fonji-ki se-mbi you(si/sin- SG)-DAT ask-OPT say(AUX)-IMPF Tv-ki sa- “to want/wish to do smth.;” “(I) want to ask you” (ZAKH:131).
gorel-1.p65
189
4/17/02, 8:57 AM
190
part four
In Manchu, many nouns are changed into syntactic words, mostly postpositions, by the attachment of the dative marker. The nouns themselves show a tendency to lose their lexical meaning, becoming partly desemanticized. Some of these words can serve as conjunctions in complex sentences. The following postpositions occur quite frequently: bade “if,” “in the case that” (ba “place,” “occasion,” “reason,” “condition”); dade “in addition to” (da “root,” “base,” “foundation”); erinde “when” (erin “time”); fonde “when” (fon “time”); jakade “to the presence of,” “up to,” “by,” “in front of” (jaka “thing,” “object”); jalinde “for the sake of” (jalin “reason,” “motive”); sidende “in between” (siden “space,” “interval”); turgunde “because,” “since” (turgun “reason,” “motive”). Syntactic functions of these words will be described in more detail in chapters dedicated to postpositions and conjunctions (see 8). Finally, if there are several homogeneous parts of the sentence to be indicated by the dative marker, only the last one is actually marked. The marker for the ABLATIVE is -ci. The ablative is grouped together with the so-called locative cases which are used to express space relationships. The ablative is used to indicate adverbial modifiers which denote the starting point in space or in time. In accordance with this basic function, the ablative displays the following syntactic meanings: 1) it indicates a starting point in space: boo-ci tuci-ke house-ABL go.away-PART “(Someone) went away from the house “ (ZAKH: 135); uba-ci goro akå this.place-ABL far COP.NEG (there.is.not) “not far from this place (from here)” (PASH2:22); abka ci wasi-mbi sky/heaven ABL descend-IMPF “(Smth.) is descending from the sky;” na ci banji-mbi earth ABL be.born-IMPF “(Smth.) is growing from the earth” (PASH1:61); bi goro ba-ci ulin bota-mbi seme ji-he I distant place-ABL wealth catch-IMP in.order.(to) come-PART
gorel-1.p65
190
4/17/02, 8:57 AM
191
morphology
bota—(ma. buta-) “to catch (game or fish);” “I came from a distant place to gain wealth;” ere ejen’i yamun-ci goro akå emu amba luce this khan(GEN) palace-ABL far there.is.not one big tower bisi-re bou (ma. boo) bi-he bi be-PART house be-PART COP “Not far from the khan’s palace there was a house with a big tower” (SK:48); 2) a starting point in time: dergi de guri-he ci east DAT move-PART ABL “since (they) moved to the east” (ZAKH:135); ajigen-ci ere-i adali ferguecuke gasha be childhood-ABL this-GEN like wonderful bird ACC àue (ma. àuwe) sabu-rakå not.at.all see-PART(NEG) “I haven’t seen a bird as wonderful as this since childhood” (SK:48); 3) a starting point of a length of road or space of time in constructions of the type “from ... to,” “from ... until”: gemun hecen ci golo-i hecen de capital city ABL province-GEN city DAT sandalabu-ha-ngge goro be.distant.from-PART-NR far
“It is a great distance from the capital city to a provincial city”
(ZAKH:135);
da-ci dube-de isitala beginning-ABL end-DAT until “from the beginning to the end” (PASH1:131); 4) the meaning of separation: ama father inenggi day
gorel-1.p65
eniye ci fakca-fi, boo-ci alja-fi mother ABL separate-CONV house-ABL leave-CONV goida-ha last.for.a.long.time-PART
191
4/25/02, 12:24 PM
192
part four
“Since he separated from (his) father and mother and left the family, a long time has passed” (ZAKH:135). In his Grammar, Möllendorff gave a different translation of the above sentence: “Taking leave of his father and mother, he was long separated from his family” (Möllendorff, 1892:6). 5) the ablative marks an object with which something is compared (an object of comparison): ere erin ci oyonggo ningge akå this time ABL important SBSTR COP.NEG (there.is.not) “There is no time more important than the present” (ORL:156); tere ere-ci sain that this-ABL good “that is better than this” (ZAKH:135); nenehe ci beye umuài labdu sain o-ho former ABL myself very much good become-PART “In comparison with the former I became much better” (SK:48); 6) the ablative has distributive meaning: ya-ci neneme ji-he bi-he what-ABL the.first come-PART be-PART “Which came first?” (MOLL:6). In Manchu, the ablative is governe d by the following verbs: fakca“to leave,” “to separate;” alja- “to leave;” tuci- “to come out,” “to go out,” “to emerge,” “to appear:” bi boo-ci je-fi tuci-ke I house-ABL eat-CONV go.out-PART “Having eaten, I went out from the house” (ORL:155); idu ci hoko-mbi, niyaman ci fakca-mbi, boo a.turn.at.duty ABL part-IMPF relative ABL separate-IMPF house ci alja-mbi, jobolon ci jaila-mbi ABL leave-IMPF disaster ABL avoid-IMPF “(Someone) leaves his turn at duty, parts from his relatives, leaves his home and avoids disaster” (PASH2:35). The postposition tulgiyen “besides” also governs the ablative:
gorel-1.p65
192
4/17/02, 8:57 AM
193
morphology tere-ci tulgiyen buye-rakå this-ABL besides desire-PART(NEG)
“(Someone) desires nothing besides (more than) this” (GAB).
As for the deri-form in classical Manchu, scholars who consider it
case marker differ in specifying its meaning. Those who recognize
deri as a case often use different terms to designate it.
According to Zakharov, the marker deri can be used instead of the marker of the ablative ci (Zakharov, 1879:136). He gives some examples in his Grammar: dorgideri “from inside;” tulergideri “from outside;” giyamun deri gene-he station from go.away-PART “(Someone) went away from (left) a station;” muke-i jugån deri ji-he water-GEN way by come-PART “(Someone) came by water” (ZAKH:136). Here are some more examples: muke gisari deri tuci-mbi wather spring from go.out-IMPF “Water stems from a spring;” edun sangga deri dosi-mbi wind opening through enter-IMPF “Wind goes through a slot;” fa deri gala be jafa-mbi window from hand ACC seize-IMPF “(Someone) from a window seizes the hand” (PASH2:65). G. Kiyose used the term “prolative” to designate the deri-case form (Kiyose, 1997:150). J. Norman classified the deri-form together with the ci-form as the ablative for classical Manchu (Norman, 1974:165). As far as Sibe is concerned with respect to these two locative cases, the distribution of meaning between the case markers ci and deri suggested by Norman is different from the meaning these markers have in Radlov’s texts. There ci is used to denote one of the functional variants of the ablative, with the main meaning of a starting point (in space or in time), and deri corresponds roughly to the separative
gorel-1.p65
193
4/17/02, 8:57 AM
194
part four
case (Lebedeva & Gorelova, 1994:49). There are a number of instances where deri is used in comparative constructions: encu hehe-ài (ma. hehe-si) deri fulu tua-mbi (ma. tuwa-mbi) other woman-PL from better consider-IMPF “(He) began to consider her better than other women;” àini (ma. sini) hojihån (ma. hojihon) tere ahån you son-in-law that elder.brother deo ningun (ma. ninggun) n’alma (ma. niyalma) deri younger.brother six man from gethun (ma. getuhun) mergen k’angk’an (ma. kiyangkiyan batur bi-me awake wise powerful hero be-CONV haha sain man good ahån deo “brothers” “Your son-in-law in comparison with his six brothers is a very wise and powerful hero and a good man” (SK:49). J. Norman refers to the deri-case form as the ablative, which does not seem quite appropriate, as the Latin term “ablative” means both “the starting point (in space or in time)” and “separation (specification)”. The ci-case form is described by Norman as directional, i.e. as opposite in its semantics to the meaning of the starting point (denoting “to” rather than “from”) (Norman, 1974:1656). 2.9. Word Formation of Nouns Most nouns are derivative, with the exception of a few non-derived ones which were mentioned above. In classical Manchu there were many productive noun forming suffixes. Nouns that have abstract meaning can be formed from verb stems by means of the following suffixes: 1) -n: acan “meeting,” “juncture,” “harmony,” “concord,” “union” (aca- “to meet,” “to get together,” “to combine,” “to be in harmony,” “to be in agreement”); buyenin “desire” (buye- “to desire,” “to admire,” “to like,” “to love,” “to do gladly”); edun “wind” (edu- “to blow” (of the wind); efin, efiyen “game,” “play” (efi, efiye- “to play”); gånin “thought,” “opinion,” “feeling,” “mind,” “spirit” (gåni- “to think,”
gorel-1.p65
194
4/17/02, 8:57 AM
195
morphology
“to consider,” “to intend”); jabun “answer” ( jabu- “to answer,” “to respond”); jodon “weave,” “grass linen” ( jodo- “to weave”); isan “assembly,” “gathering,” “a meeting” (isa- “to assemble,” “to come together,” “to gather”); tacin “learning,” “science,” “skill” (taci- “to teach”). With regard to these nouns it should be mentioned that in some cases it is unclear whether the nouns derived from the verb or whether the latter is of nominal derivation. 2) the suffix -n is used to form nouns from verbs which contain the suffix used for forming the passive and causative moods, e.g. -bu: alibun “a petition” (alibu- “to present a document to a superior”); banjibun “a compendium,” “a creation,” “a product” (banjibu- “to give birth to,” “to compile,” “to compose,” “to make up,” “to form”); ejebun “record,” “notes” (ejebu- “to be recorded,” “to be noted down”); hesebun “fate,” “determination” (hesebu- “to ordain,” “to determine”); tuwabun “survey,” “review,” “view” (tuwabu- “to be on view,” “to show,” “to exhibit”); ulabun “tradition,” “what is handed down,” “biography” (ula- “to hand down,” “to pass on,” “to hand on,” “to pass to;” ulabu- PASS/CAUS of ula-); yendebun “inspiration,” “excitement,” “inducement” (yendebu- “to induce,” “to inspire”); 3) -lan/-lon/-len , -ran/-ron/-ren which can be analyzed historically as composed of suffixes -la/-lo/-le, -ra/-ro/-ra plus -n; the suffixes -la/ -lo/-le and -ra/-ro/-re are in face verbal ones. Both non-derived and derivative verbs contain them in their morphological structure and it would be accurate to consider only the suffix -n as derivative for the majority of such nouns: aisilan “relief,” “aid,” “assistance” (aisila“to help,” “to aid,” “to assist,” “to provide”); dorolon “rite,” “cere mony” (dorolo- “to perform a rite”); cihalan “desire,” “wish” (cihala“to like,” “to be fond of,” “to wish”); gisuren “talk,” “discussion” (gisure“to speak,” “to talk”); jobolon “trouble,” “disaster,” “sorrow,” “mourn ing” (jobo- “to suffer,” “to be in need,” “to worry,” “to be distressed”); ergelen “force,” “coercion” (ergele- “to force,” “to coerce”); erulen “tor ture,” “torment” (erule- “to torture,” “to punish”); tacuran “com mission,” “duty,” “mission” (tacura- “to send on a mission,” “to dele gate,” “to commission”); wakalan “fault” (wakala- “to blame,” “to fault,” “to accuse,” “to impeach,” “to deem wrong”); 4) -kiyan/-kiyen, -hiyan/-hiyen which can also be analyzed historically as composed of suffixes -kiya/-kiye, -hiya/-hiye plus -n which is, to be
gorel-1.p65
195
4/17/02, 8:57 AM
196
part four
precise, the only derivative one for this group of nouns. The rest belong to the verbal stems: dasihiyan “dusting,” “shaking off” (dasihiya“to dust smth.,” “to dust with a feather duster,” “to shake off”); necihiyen “calming,” “soothing,” “pacification;” “peaceful,” “tranquil,” “serene” (necihiye- “to console,” “to calm down,” “to pacify,” “to sub jugate”); tacihiyan “teaching,” “training” (tacihiya- “to instruct,” “to teach”); tuwakiyan “supervision,” “safeguarding,” “guarding (from, against);” “discretion in conduct” (tuwakiya- “to supervise,” “to watch,” “to guard,” “to watch over,” “to observe”); 5) -gan/-gen/-gon, -han/-hen/-hon: aligan “support,” “retainer,” “base” (ali- “to receive,” “to accept,” “to support,” “to hold up”); bodogon “consideration,” “plan” (bodo- “to plan,” “to think over”); dengnehen “balance,” “scales;” “the leg of a boot” (NL:58) (dengne- “to weigh on a small steelyard,” “to balance,” “to complete with”); joligan “ran som money,” “ransom” ( joli- “to redeem,” “to ransom”); kubuhen “border,” “edging,” “hem” (kubu- “to add a border to,” “to edge,” “to hem”); nemehen “addition,” “increment” (neme- “to add,” “to increase”); nirugan “picture,” “chart,” “map,” “drawing,” “painting” (niru- “to draw,” “to paint”); sujahan “supporting pole,” “prop,” “support” (suja- “to prop up,” “to support”); 6) the suffix -ku/-kå is used to form verbal nouns which denote instruments and utensils, occupation, profession and rank: alikå “a tray,” “the tray used for weighing on a scale” (ali- “to receive,” “to except;” “to support,” “to hold up”); anakå “key” (ana- “to push”); anjikå “hatchet” (anji- “to hack,” “to chop with a hatchet”); berileku “a drill” (berile- “to drill”); cirgeku “a wooden implement used to pound earth,” “a ramrod” (cirge- “to ram,” “to pound earth”); dabukå “fire lighter,” “small stove,” “burner” (dabu- “to light a fire or lamp”); dengneku “a small steelyard” (dengne- “to weigh on a small steelyard”); dobokå “sacrificial candlestick” (dobo- “make a sacrifice,” “to offer up,” “to offer in a ceremony”); etuku “clothing,” “garment” (etu- “to put on clothing,” “to wear”); harikå “an iron (for pressing clothing),” “a cauterizing iron” (hari- “to iron,” “to press (clothing),” “to cauterize sores on cattle”); haàakå “a scrubbing brush made from kaoliang stalks or the stalks of other grains” (haàa- “to scrub”); hedereku “a rake” (hedere- “to rake”); korikå “chisel,” “gouge,” “a small curved knife” (kori- “to hollow out,” “to dig out,” “to cut out”); laidakå “calumniator,” “slanderer,” “mischievous,” “ill-behaved,” “self-willed
gorel-1.p65
196
4/17/02, 8:57 AM
197
morphology
children” (laida- “to calumniate,” “to slander,” “to blame someone else for one’s own errors,” “to be mischievous,” “to be self willed”); nereku “a sleeveless rain cape made from leather or oilskin” (nere- “to throw over the shoulders (clothing);” niyeleku “a stone roller,” “upper millstone,” “a stick for washing” (niyele- “to roll,” “to mill (grains),” “to roll fabrics with a stone roller”); àokå “a currycomb” (ào- “to scrape,” “to scrape off,” “to curry” (livestock); tacikå “school” (taci“to learn,” “to study”); 7) the suffix -ku/-kå is used to form nouns from verbs that have the suffix -bu (the form of the passive and causative moods). This group of verbal nouns has several meanings: instruments of human performance; a person who is characterized according to the action expressed by a verb from which this noun is derived; professions: acabukå “flatterer,” “an obsequious person” (acabu- “to be obsequious,” “to flatter”); hafumbukå “interpreter,” “translator” (hafumbu- “to translate,” “to interpret”); tacibukå “teacher,” “instructor” (tacibu- “to teach,” “to instruct”); teherebuku “balance,” “scales” (teherebu- “to to weigh on a balance”); ubaliyambukå “translator” (ubaliyambu- “to translate from one language into another”); 8) the suffix -cun forms verbal nouns of various abstract meanings: akacun “sadness,” “grief” (aka- “to be sad,” “to grieve”); akdacun “trust,” “trustworthiness,” “what one depends on” (akda- “to depend on,” “to trust”); aliyacun “regret,” “waiting” (aliya- “to regret,” “to wait”); elecun “satiety,” “satisfaction,” “content,” “pleased,” “satisfied” (ele“to suffice,” “to be enough”); gasacun “grudge,” “complaint” (gasa“to grieve,” “to complain,” “to hold a grudge”); girucun “shame,” “disgrace,” “disgraced” (giru- “to be ashamed,” “to feel ashamed,” “to be embarrassed”); isecun “apprehension,” “fear,” “frightened,” “overawed” (ise- “to apprehend,” “to fear,” “to lack courage,” “to be timid”); korsocun “annoyance,” “regret” (korso- “to regret,” “to miss,” “to be annoyed at,” “to hate”); suilacun “distress,” “labour,” “agony,” “hardship” (suila- “to be in distress,” “to suffer hardship,” “to be exhausted,” “to be in agony”); ulhicun “understanding,” “insight,” “knowledge” (ulhi- “to understand,” “to comprehend”); yertecun “shame” (yerte- “to be ashamed,” “to be embarrassed”); 9) -sun, -fun: adasun “lap,” “lapel” (ada- “to accompany,” “to be attached to,” “to stitch together”); hadufun “scythe,” “sickle” (hadu
gorel-1.p65
197
4/17/02, 8:57 AM
198
part four
“to cut with a sickle,” “to reap”); håwaliyasun “harmony,” “concord,” “union,” “harmonious,” “in concord,” “united” (håwaliya- “to har monize,” “to unite,” “to reconcile,” “to conciliate”); ijifun “comb” (iji- “to comb,” “to put the vertical threads on a loom,” “to put in proper order”); umiyesun “girdle,” “girth,” “belt,” “sash” (umiyele- “to tie a belt,” “to gird oneself”); 10) the suffixes -si, -msi, -ci, -ji, -lji, -mji, -nju are used to form nouns both from verbs and nouns. The resulting nouns designate people according to their skills, occupations, ranks, trades, and business: adulasi “herdman” (adula- “to herd”); usisi “farmer,” “cultivator of the land” (usin “field for cultivation”); bithesi “scribe,” “secretary,” “clerk” (bithe “book,” “letter”); kumusi “musician” (kumun “music”); giyamusi “inspector of a post station” (giyamun “relay station,” “relay post,” military post station”); sejesi “driver,” “coachman” (sejen “cart”); ulandusi “postman,” “letter-carrier,” “Superintendant of a Military Post” (ulandu- “to hand down from one person to another,” “to pass on from one person to another”); kimcisi “investigator” (kimci- “to examine,” “to check,” “to investigate,” “to look into carefully”); tacimsi “student of the Imperial Academy of Learning” (taci- “to learn,” “to study,” “to be accustomed”); dahalasi “follower” (dahala- “to follow,” “to pursue,” “to run down”); tuwakiyasi “guard,” “watchman” (tuwakiya- “to guard,” “to watch,” “to watch over,” “to observe”); beidesi “judge” (bejde- “to examine a case,” “to try a case,” “to judge”); jubesi “storyteller,” “taleteller” (juben “story,” “tale”); aduci “herder” (adun “herd,” “swarm”); morici “groom,” “stableman,” “a person who watches horses at official establishments” (morin “horse”); sejeci “cart maker,” “person in charge of official vehicles” (sejen “cart,” “vehicle”); boigoji “host,” “master” (boigon “family,” “household,” “property”); dahalji “a manservant under a life contract” (dahala- “to follow,” “to pursue,” “to run down”); butemji, butumji “cunning,” “deceitful” (butu “dark,” “dim,” “hidden,” “secret;” butule- “to cover,” “to cover up,” “to act secretly;” butu- “to hibernate”); ukanju “fugitive” (uka- “to to flee,” “to run away,” “to desert”). In some cases the suffix -ci attached to nouns, may form nouns that have the meaning of objects which are obtained as a result of doing a trade, a business, a handicraft: ihaci “cowhide” (ihan “bovine,” “cow,” “ox,” “bull”); honci “sheepskin” (honin “sheep”); nimaci “goat skin” (niman “goat”); nuheci “hide of the wild boar” (nuhen “a oneyear-old wild pig”) (Zakharov, 1879:74);
gorel-1.p65
198
4/17/02, 8:57 AM
199
morphology
11) the suffix -tun is used to form nouns from both nouns and verbs. The resulting nouns denote mainly instruments and utensils: tetun “tool,” “implement,” “vessel,” “dishes” (tebu- “to pour,” “to fill a vessel,” “to pack,” “to put in”); alitun “a cup,” “a small table for offerings” (ali- “to receive,” “to accept,” “to support,” “to hold up”). In Manchu, a number of suffixes are used to form a group of nouns with semantics of quality, which is one of the properties of object nouns. These nouns are associated with adjectives and translated by them. They may also denote a person with regard to certain qualitative characteristics ascribed to him/her, and in that case they are treated as proper nouns. The following suffixes form nouns both from nouns and verbs: 1) -tu, -du: albatu “tributary,” “rough,” “rude,” “rustic,” “common,” “ordinary” (alban “public service,” “official,” “public,” “fiscal,” “tax,” “duty,” “tribute”); bekitu “strong man,” “strong,” “stable” (beki “strong,” “stable,” “firm”); dahaltu “orderly,” “manservant” (daha- “to follow,” “to submit,” “to surrender,” “to obey”); giratu “big-boned (of livestock)” (giran “corpse;” giranggi “bone”); girutu “ashamed,” “embarrassed” (giru- “to be ashamed,” “to feel ashamed,” “to be embarrassed”); hålhatu “thief,” “swindler,” “thievish” (hålha “thief,” “robber,” “secret”); hålhitu “a muddleheaded person” (hålhi “muddle headed,” “confused,” “blurred”); niyakitu “a dirty-nosed child” (niyaki “pus,” “nasal and bodily discharge”); songgotu “a person who cries frequently,” “a crybaby” (songgo- “to cry,” “to weep”); turgatu “a skinny person” (turga “thin,” “skinny,” “lean”); urhåtu “easily frightened (of horses and other livestock)” (urhå- “to shy” (of livestock); urhån “shyness” (of livestock); urhutu “leaning to one side,” “having one leg shorter than the other,” “a person who leans to one side” (urhu “onesided,” “tilting,” “partial;” urhu- “to lean to one side,” “to be partial,” “to be prejudiced to one side”); 2) -ta, -da, -to, -do: alhata “variegation,” “diversity of colours,” “mottled patterns,” “variegated patterns” (alha “many coloured,” “variegated,” “mottled”); fulata “red-eyed,” “having red circles about the eyes” (fula hån “pink,” “reddish”); ganggata “giant,” “tall in stature” (gangga, ganggan “hard,” “strong,” “tough”); giohoto “beggar” (gioha-, gioho- “to beg” (for alms); kaikata, kaikada “squint-eyed/cross-eyed person,” “slanting,” “askance” (kaikara- “to be slanted,” “to be crooked,” “to look at askance”); manda “slow,” “late;” sanggata “having holes or
gorel-1.p65
199
4/17/02, 8:57 AM
200
part four
openings” (sangga “hole,” “opening”); suwaliyata “mixture,” “blend,” “mixed,” “blended” (suwaliya- “to mix,” “to mix up,” “to blend together,” “to combine,” “to put together;” suwaliyan “mixture,” “mixing”); ubiyada “detestable,” “hateful,” “execrable,” “loathsome” (ubiya- “to detest,” “to loathe”); 3) -ki: acabuki “flatterer,” “sycophant” (acabu- “to be obsequious,” “to flatter”); adaki “neighbour,” “neighboring,” “adjacent” (ada- “to be close to,” “to be next,” “to be attached to”); ambaki “haughty,” “proud,” “pompous” (amba “big,” “great,” “vast,” “important”); goroki “distant,” “distant place,” “distance” (goro “far,” “distant”); hanciki “near,” “near place,” “vicinity” (hanci “near,” “close”); horoki “having a senile aspect;” jabàaki “a good fortune,” “advantage,” “luck,” “lucky” (jabàa- “to obtain an advantage,” “to derive benefit from,” “to be by good luck”); juseki “childish,” “juvenile,” “childhood” (juse PL of jui “child,” “son”); oyoki “hasty,” “hurried,” “bustler” (oyo- “to be more than half finished,” “to be almost done”); sakdaki “having the aspect of old age (sakda “old,” “aged,” “old man”); ufaraki “a slight error” (ufara- “to err,” “to fail,” “to make a mistake about something,” “to fail,” “to lose interest”).
3. Numerals In the Manchu language there are several groups of numerals characterized by specific semantics and distinctive morphological markers (suffixes). The cardinal numerals are the morphological base for other groups of numerals. These are ordinal, distributive, multiplicative, and fractional numerals. Used without nouns, the cardinal and ordinal numerals may be declined by being followed by case markers: emu be sa-me mute-he de, juwe be one ACC know-CONV be.able-PART DAT two ACC sa-ra-ngge mangga akå know-PART-NR difficult COP.NEG (there.is.not) “If (someone) was able to know “one,” it is not difficult to know “two”(ZAKH:107). When the cardinal and ordinal numerals are followed by nouns, the latter take case markers:
gorel-1.p65
200
4/17/02, 8:57 AM
201
morphology ilaci jui be uba-de ji-bu-kini the.third son ACC this.place-DAT come-CAUS-OPT “Let the third son come here (at this place)” (ZAKH:107). 3.1. Cardinal Numerals
The Manchus used the standard decimal positional system. For the first ten numbers the cardinal numerals are: emu “one,” juwe “two,” ilan “three,” duin “four,” sunja “five,” ninggun “six,” nadan “seven,” jakån “eight,” uyun “nine,” juwan “ten.” Numerals for the second group of ten numbers, with one exception, are composite. They are formed by adding one of the cardinal numerals of the first ten to the word juwan “ten:” juwan emu “eleven” (ten and one); juwan juwe “twelve” (ten and two); juwan ilan “thirteen” (ten and three), juwan duin “fourteen” (ten and four), etc. Only for the numeral “fifteen” a special word tofohon is used. Cardinal numerals denoting multiples of ten in the first hundred are: orin “twenty,” gåsin “thirty,” dehi “forty,” susai “fifty,” ninju “sixty,” nadanju “seventy,” jakånju “eighty,” uyunju “ninety.” In Manchu there are also special words for the following powers of ten: tanggå “hundred;” minggan “thousand;” tumen “ten thousand.” For some larger powers of ten the following composite words are used: juwan tumen “a hundred thousand,” tanggå tumen “a thousand thousand (million),” minggan tumen “ten million,” and tumen tumen “a hundred million.” Equally with these composite words, the loan cardinal numerals borrowed from Sanskrit were also used: bunai, bujun— juwan tumen “a hundred thousand,” saja—tanggå tumen “a thousand thousand (million),” jirun—minggan tumen “ten million,” dungàun—tumen tumen “one hundred million.” For greater powers of ten, starting from a thousand million, the Manchus used loan words borrowed from Sanskrit at the time when Buddhist works were translated. These numerals are: terbun (10 dungàun) – “a thousand million,” “billion;” cakcin (10 terbun) – “ten thousand million,” “ten billion;” jabsun (10 cakcin) – “a hundred thousand million,” “a hundred billion;” damdan (10 jabsun) – “a trillion;” jiri (10 damdan) – “ten trillion;” miàun (10 jiri) – “one hundred trillion;”
gorel-1.p65
201
4/17/02, 8:57 AM
202
part four
cata (10 miàun) – “a quadrillion;”
ganggi (10 cata) – “ten quadrillion;”
jamuri (10 ganggi) – “one hundred quadrillion;”
samuri (10 jamuri) – “one sextillion;”
samina (10 samuri) – “one ten-sextillion;”
buju baja (10 samina) – “innumerable,” “uncountable,” “countless
numbers.” Numbers expressed by several figures are arranged in such a way that the largest figures are placed in the initial position, and the others are listed one after another according to their quantities, for instance: ilan minggan hadan tanggå orin duin “three thousand seven hundred and twenty four;” emu minggan jakån tanggå nadanju ninggun “ one thousand eight hundred and seventy six.” When a multitude of objects are indicated by numbers, the cardinal numerals are normally placed before nouns denoting the quantity of objects: ilan niyalma ji-he three man come-PART “Three men came” (ZAKH:89). When counting objects, the cardinal numerals are normally placed after the noun denoting objects which are numerated: honin juwan “ten sheep,” ihan duin “four bulls,” morin sunja “five horses.” Instead of the cardinal numeral emu “one,” especially in spoken language, the cardinals emke or emken are used. Taking the marker of the dative only or adding the word oci (originally the form of the conditional converb from the verb o- “to be,” “to become” which is partly desemanticized in this position), the cardinal numerals may be the base for forming adverbs: emu de/ emu de oci “in the first place;” jaide or jaide oci “in the second place;” sunja de or sunja de oci “in the fifth place.” Followed by the words sasa, emgi “together,” the cardinal numerals also may serve as the base for forming adverbs: ilan hofi sasa jura-ka three person together set.out-PART “Three (together) set out” (ZAKH:103). Cardinal numerals may be followed by the suffix -ngga/-ngge/-nggo: duingge “folded in four,” “in fours;” sunjangga “pertaining to the
gorel-1.p65
202
4/17/02, 8:57 AM
203
morphology
number five;” sunjangga inenggi “the fifth day of the fifth month (the Dragon Boat Festival);” nadangga “pertaining to the number seven;” nadangga inenggi “the seventh day of the seventh month;” uyungge “pertaining to the number nine;” uyungge inenggi “the ninth day of the ninth month” (Zakharov, 1879:107; NL). 3.2. Ordinal Numerals Ordinal numerals are formed by adding the suffix -ci to the cardinal numbers. Final -n is deleted before this suffix, with three exceptions to the rule, which are juwanci “the tenth,” mingganci “the thousandth” and tumenci “the ten thousandth.” There are several variants for “the first,” namely: uju (originally, “head”), ujui (uju “head” + i = GEN), ujuci, emuci. There are also several variants for “the second” which are jai, jaici, juweci. The ordinal numerals uju, ujui “the first” and jai “the second” are used to numerate the numbers of the first ten. To numerate the numbers of the second and the following tens the regular variants emuci and juweci are used: juwan emuci “the eleventh;” orin juweci “the twenty second.” Instead of emuci, the words bonggo “the first,” “number one,” “leading” and sucungga “the first,” “initial,” “beginning” are used to indicate years of an emperor’s reign. Other ordinal numerals are: ilaci “the third,” duici “the fourth,” sunjaci “the fifth,” ningguci “the sixth,” nadaci “the seventh,” jakåci “the eighth,” uyuci “the ninth,” juwanci “the tenth,” tofohoci “the fifteenth,” orici “the twentieth,” gåsici “the thirtieth,” dehici “the fortieth,” susaici “the fiftieth,” ninjuci “the sixtieth,” nadanjuci “the seventieth,” jakånjuci “the eightieth,” uyunjuci “the ninetieth,” tanggåci “the hundredth,” mingganci “the thousandth,” tumenci “the ten thousandth.” In ordinal numerals, expressed by several figures, the suffix -ci is added only to the last word: juwan emuci “the eleventh,” juwan juweci “the twelfth,” juwan jilaci “the thirteenth,” orin ilaci “the twenty third,” ninju jakåci “the sixty eighth,” emu minggan jakån tanggå nadanju ningguci “the one thousand eight hundred and seventy sixth.” Ordinal numerals may be followed by the suffix -ngge which can be described here as a nominalizer. In these cases, ordinal numerals have the same function as substantives: sunjaci-ngge de goi-ha the.fifth-NR DAT become-PART “(Someone) became number five in the examination;”
gorel-1.p65
203
4/17/02, 8:57 AM
204
part four
duici-ngge be gai-ha the.fourth-NR ACC marry(a wife)-PART “(He) married the fourth (wife)” (ZAKH:107). 3.3. Distributive Numerals Distributive numerals are formed by adding the suffix -ta/-te/-to to the cardinal numbers. The final -n is deleted before this suffix, the exception to the rule being juwan “ten.” However, in an earlier stage of the language, the variant juwata was also possible. The distributive numerals are: emte “one each,” juwete “two each,” ilata “three each,” duite “four each,” sunjata “ five each,” ninggute “six each,” nadata “seven each,” jakåta “eight each,” uyute “nine each,” juwata (juwanta) “ten each,” tofohoto “fifteen each,” orita (orinta) “twenty each,” gåsita “thirty each,” dehite “forty each,” susaita “fifty each,” ninjute “sixty each,” nadanjute “seventy each,” jakånjute “eighty each,” uyunjute “ninety each,” tanggåta “one hundred each,” minggata (emte minggan) “one thousand each,” tumete (emte tumen) “ten thousand each.” The harmonical variants ningguta, dehita and nadanjuta were mostly used in old Manchu (Zakharov, 1897:105). The distributive numerals may also be formed by repeating the word denoting a number followed by the genitive marker: emke emken i “one by one,” ilan “three”—ilan ilan i “by threes;” juwan “ten”— juwan juwan i “by tens.” In distributive numerals, expressed by several figures, only the last figure takes the suffix -te, for instance: juwan duite “fourteen each.” The distributive numerals including the words minggan “thousand” and tumen “ten thousand,” may be expressed in two ways: minggata or emte minggan “one thousand each;” tumete or emte tumen “ten thousands each” (Zakharov, 1879:1034; Pashkov, 1963:28-9; NL). 3.4. Fractional Numerals The fractional numerals are formed by the word-combination the first word of which is a cardinal numeral followed by the ablative marker -ci. This is how a denominator is indicated. The second word of the word-combination is a cardinal numeral indicating the number of parts, that is the numerator: ilan ci emu “a third,” duin ci emu “a fourth,” sunja ci ilan “three fifths,” nadan ci juwe “two sevenths.” Between two cardinal numerals the word ubu “part” followed by
gorel-1.p65
204
4/17/02, 8:57 AM
205
morphology
the genitive (marker i) may be inserted: duin ubu i emu “a fourth,” sunja ubu i ilan “three fifths,” nadan ubu i juwe “two sevenths” (Pashkov, 1963:29). 3.5. Multiplicative Numerals Multiplicative numerals are formed by adding the suffix -rsu to the cardinal numbers and dropping the final -n. There are two exceptions, i.e. juwan and tumen where the final -n is not deleted. The cardinal numeral juwe “two” loses the last syllable -we. Multiplicatives may be also formed by addition of the word ubu “part” to the cardinals. Here are some examples: emursu, emu ubu “single;” jursu, juwe ubu “double;” ilarsu, ilan ubu “threefold;” uyursu, uyun ubu “ninefold;” tanggårsu “one hundredfold”4. With reference to textile fabrics the suffix -ri is used attached to the cardinal numbers: ilari “threefold,” sunjari “fivefold,” jakåri “eightfold” (Möllendorff, 1892:8). 3.6. Collective Numerals The suffix for the collective numerals is -nofi. The final -n of the cardinal numerals to which it is added, is normally deleted: juwenofi “two (together),” ilanofi “three (together),” duinofi “four (together),” sunjanofi “five (together),” etc.5. 3.7. Iterative Numerals Iterative numerals are normally formed by adding the suffix -nggeri to the cardinal numbers if they end in a vowel, or the suffix -geri if the cardinals end in a consonant. In this case the final -n changes into -ng: emgeri (instead of emugeri) “once,” juwenggeri “twice,” ilanggeri “three times,” duinggeri “four times,” sunjanggeri “five times,” ninggunggeri “six times,” nadanggeri “seven times,” jakånggeri “eight times,” uyunggeri “nine times,” juwanggeri “ten times,” etc. The iterative numerals can be formed analytically when the cardinal numbers are followed by 4 In Norman’s Lexicon one can find different translations of several multiplica tive numerals: emursu “having one layer,” “simple,” “unlined;” jursu “two-layered,” “double,” “complex,” “complicated,” “pregnant;” ilarsu “three-tiered,” “three-leveled,” “three-storied” (NL:75, 166, 146). 5 The word form juwenofi is given by J.Norman as the word combination juwe nofi “two persons,” “two people” (NL:168).
gorel-1.p65
205
4/17/02, 8:57 AM
206
part four
the word mudan or meyen “time:” emu mudan “once,” juwe mudan or meyen “twice,” etc. 3.8. Numeratives In Manchu, similar to Chinese, there is a considerable number of measure words (numeratives) used with objects which are counted. Normally these words have certain lexical meanings, but when they function as numeratives they reveal the tendency to lose their meanings as lexical units and become grammatical elements. The measure words are used to indicate the quantity of objects. Depending on the semantic group of nouns by which objects are expressed, different numeratives are used. These words are placed between the cardinal numeral and the objects which are counted: ilan fesin loho “three swords” (fesin is a numerative used with objects equipped with a handle). If objects, which are being enumerated, are known from a question, then in the answer to this question a numerative can be used as a substitute for these objects. Thus, numeratives can be used anaphorically, for example, in the answer to a question where the noun is omitted: benji-he-ngge udu juru gålha deliver-PART-NR how.many? pair boot “How many pairs of boots have been delivered?” ilan juru “three pairs” (ZAKH:90). The numeratives as listed by Zakharov in his Grammar are presented in the following table. Table 6 numerative
used with nouns denoting:
afaha “sheet (of paper),” “list,” “page” aligan “support,” “base”
sheets of paper, lists of smth., registers, rolls,
inventories, bills
objects having support (pedestals, statues,
obelisks, buildings with protruding
foundation)
animals, objects having holes (pots, bags,
etc.), seaports, mountain-passes
people, persons in a household
objects forming a pair but used separately
angga “mouth,” “hole,” “pass,” “gate” anggala “household” bakcin “match,” “opposite number,” “opponent,” “the opposite side”
gorel-1.p65
206
4/17/02, 8:57 AM
207
morphology Table 6. Cont. numerative
used with nouns denoting:
baksan “a bundle”
bundles of wood, faggot, hay; sheaves (of
papers), bunches (of keys), bales, pack camels
tethered to one another, caravans, groups of
troops
beds, blankets, bases for statues of Buddha
folded papers, sheets in album, bundles of
papers or paintings
scrolls of pair inscriptions used as paintings,
folded writing-books of a kind of an album
bodies of people, carcasses of animals,
complete sets of smth.
planes or sides of a square, objects having
plate front side, viz, badges, flags, mirrors,
fans, as well as pieces of sliced meat
carrying-pole loads
objects having handle, grip; staples
parts or chapters of a book, notebooks
bound from sheets of paper
besergen “bed”
bukdan “a bend,” “a fold”
bukdarun “roll,” “scroll,” “rolled up paper” beye “body,” “self” dalgan “flatness,” “plane” damjan “a carrying pole” dasin “handle,” “grip” debtelin “notebook,” “chapter of a book,” “a volume (of an old-style book)” dedun “a post station,” “an overnight stopping place” defelinggu “bolt (of cloth)” delhen “a lot,” “part” dengjan “lamp,” “light” dobton “sheath,” “a small bag (for a seal),” “a container,” “scabbard” ergen “breath” erguwen “a period of twelve years,” “a cycle” erin “time,” “hour” farsi “piece,” “strip” fekun “a leap,” “a jump” fempi “a paper seal used on envelopes and on doors” feniyen “flock,” “herd,” “crowd” fesin “haft,” “hilt,” “shaft,” “handle” fiyelen “chapter,” “section of a book” fiyen “the feathers on the arrow shaft” fiyentehe “petal,” “clove (of garlic),” “a section,” “a slice” fulmiyen “bundle,” “package”
gorel-1.p65
207
overnight stopping places and stations
whole bolts of cloth for one or two garments
lots of ploughed land
lamps, lights, lanterns, candle-sticks
cases, instrument-cases, covers, volumes
deep breaths, breathing-spaces, respites
cycles of numeration of years
according to twelve animals
hours, dinners, suppers, hours for eating
pieces of sliced meat, rags, shreds, drugs cut
in layers—plasters
leaps, jumps, footsteps
envelopes, packets, letters
herds, flocks, runs, crowds
knives, swords, sabres, spades, shovels,
fans, umbrellas
chapters and sections of books, paragraphs
and articles of law, strophes
arrows with featherings
petals of flowers, leaves of plants,
cloves (of garlic), fragments of broken glass,
pieces of broken crockery
bundles (of wood, hay), packs, packages,
bales
4/17/02, 8:57 AM
208
part four Table 6. Cont.
numerative
used with nouns denoting:
gala “hand,” “arm”
hand sazhens equalling two Chinese feet and five inches branches, branches of a river, leaves of a door, limbs, groups of troops, parts of a whole houses and buildings situated parallel, blocks of buildings words, sentences, statements, periods of speech, verses rooms, houses
gargan “branch,” “branch of a river,” “leaf of a door” girin “strip,” “line,” “row” gisun “speach,” “word,” “language” giyalan “space between,” “interval,” “interstice” gubsu “bud (of s flower)” guksen “clouds,” “a blast of wind” hacin “kind,” “sort,” “class,” “item” hafure “going through” haha “male,” “man” hoào “corner,” “edge” hufan “company,” “troupe,” “artel” hungken “forge,” “smelting furnace” ice “new,” “the first ten days of the month,” “the first day of a lunar month” ikiri “pair,” “twins” imari “mu (a Chinese measure of land area)” inden “a rest after work,” “stopover (on a juorney)” jalan “a joint,” “generation,” “world,” “rank” jemin “a dose” jijun “stroke,” “line,” “hand (writing)” jukte “piece” juru “pair,” “doubled” justan “a strip,” “a stripe” karan “platform,” “stage” kiya “cell,” “partition” kiyan “paper measure equaling twenty
gorel-1.p65
208
buds of flowers clouds of various type various objects and things, goods, articles of law openings, holes lands granted to a man enrolled into military or government service edges, corners, sides artels of workmen or peasants, troupes of players or singers forges, smelting furnaces, quantity of copper money: 5662 strings and 369 pieces the first ten days of each month in a yeara year pair objects of various kinds, tear-off coupons or receipts quantity of arable fields or lands stopovers (on a journey), sections of plants, joints, generations doses (of medicine) strokes and lines, letters and hands, drafts long and thick pieces of meat objects forming a pair—boots, even numbers, distiches, pair-horses elongated objects of various kinds theatrical perfomances, entertainments, plays cells, honeycombs quantity of paper
4/17/02, 8:57 AM
209
morphology Table 6. Cont. numerative
used with nouns denoting:
five sheets,” “quires” mangga (instead of angga) “mouth,” “mouth of animal” mari “time,” “occasions” mudan “time,” “curve,” “sound,” “tone” meyen “part,” “section,” “segment” okson “step,” “pace” nofi “person” ufuhi “part,” “share,” “portion” uhun “bandle,” “package” ulcin “string” undehen “board,” “rod,” “plank,” “staff” ursu “layer,” “level”
animals with a big mouth, drinks (of water), mouthfuls, pots with wide openings times and occasions, turns and revolutions, methods or ways of motion and action, chapters and sections in narratives, melodies sections, segments, divisions, pieces, chapters, groups, squadrons, ranks (of troops) steps of people and animals persons having high ranks, persons (polite indication) parts, shares of various nature packages, bundles (of clothing) strings (of beads, cash, etc.) planks, wooden printing blocks, beats with a bamboo rod levels of buildings, rows of troops, objects consisting of layers (boxes, baskets, cups one smaller than the next, placed one in another
Words denoting measures of length, quantity, distance, area and weight can also be referred to as numeratives. The system of measures and weights standardized during the early Qing period by the Board of Works (ma. weilere arara fiyenten), is given in Norman’s Lexicon (NL:319-20). There are some additional measure words in Zakharov’s Grammar. They correlate to a large extent with the Russian measure terms which were in use in the author’s lifetime (Zakharov, 1879:979). 3.9. Words Denoting Periods of Time. Cycles In Manchu, specialized words were used to indicate various periods of time (ma. erguwen “a period of twelve years,” “a cycle”). These words were used to enumerate years, months, and days according to the calendar as well as chronological years. According to the
gorel-1.p65
209
4/17/02, 8:57 AM
210
part four
tradition of the Oriental World, the Manchus used three Cycles. They are the10-element Cycle, the 12-element Cycle and the 60-element Chronological Cycle. The 10-element Cycle, in Manchu juwan cikten “the ten heavenly stems,” was used to enumerate objects up to ten and days in each decade of the month. But above all it was used to numerate chronological years combining its terms with those of the 12-element Cycle. The terms of the 10-element Cycle were named according to the ten heavenly stems or colors of the five elements, namely, earth, air, fire, water, and metal. The 10-element Cycle is represented by the following terms:
1) niowanggiyan “green (the colour of tree),” “the first of the heavenly
stems”6
2) niohon “greenish,” “the second of the heavenly stems”
3) fulgiyan “red (the color of fire),” “the third of the heavenly stem”
4) fulahån “reddish,” “the fourth of the heavenly stems”
5) suwayan “yellow (the colour of the earth),” “the fifth of the heavenly
stems”
6) sohon “yellowish,” “the sixth of the heavenly stems”7
7) àanggiyan “white (the color of metal);” “the seventh of the heavenly
stems”
8) àahån “whitish,” “the eighth of the heavenly stems”
9) sahaliyan “black (the colour of water),” “the ninth of the heavenly
stems”
10) sahahån “blackish,” “the tenth of the heavenly stems”
The 12-element Cycle, in Manchu juwan juwe gargan “the twelve
earthly branches,” was used to enumerate objects up to twelve, for
instance, signs of the zodiac, hence, months of the year, the year of
one’s birth, calendar years, hours of the twenty-four hour period,
etc. The terms of this Cycle were given according to the twelve names
of animals.
6 The word niowanggiyan is translated as “ninth of the earth’s branches” by Norman (NL:215). 7 The word sohon means “deep yellow (the color of the sunflower)” according to Norman (NL:247).
gorel-1.p65
210
4/17/02, 8:57 AM
211
morphology The 12-element Cycle is represented by the following terms:
1) singgeri “rat,” “mouse,” “the first of the earthly branches”
2) ihan “bull,” “the second of the earthly branches”
3) tasha “tiger,” “the third of the earthly branches”
4) gålmahån “hare,” “the fourth of the earthly branches”
5) muduri “dragon,” “the fifth of the earthly branches”
6) meihe “snake,” “the sixth of the earthly branches”
7) morin “horse,” “the seventh of the earthly branches”
8) honin “sheep,” “the eighth of the earthly branches”
9) bonio “monkey,” “the ninth of the earthly branches”
10) coko “chicken,” “the tenth of the earthly branches”
11) indahån “dog,” “the eleventh of the earthly branches”
12) ulgiyan “swine,” “pig,” “the twelfth of the earthly branches”
The 60-element Cycle is used to enumerate chronological years in
the History of the World. It is formed by combination of a term from
the 10-element Cycle with a term from the 12-element Cycle, i.e.
the first element for the 60-element Cycle is niowanhhiyan singgeri
“green mouse”, etc., up to the tenth term of the first Cycle. To create
the eleventh element of the 60-element Cycle, the first term of the
10-element Cycle and the eleventh term of the 12-element Cycle are
used. To create the thirteenth element of the 60-element Cycle, the
third term of the 10-element Cycle is combined with the first term
of the 12-element Cycle, etc. When all the terms of the 10-element
Cycle have been listed the numeration starts again, and its first term
combines with those terms of the 12-element Cycle which have not
yet been combined with a term of the first Cycle. In addition, when
all the terms of the second Cycle are over then the first of its terms
combines with an unlisted term of the first Cycle. Thus, each term
of the first Cycle is combined with each term of the second Cycle
six times, but any term of the second Cycle is combined with any
term of the first Cycle only five times. The first element of the next
60-element Cycle, is again formed by the first term of the first
Cycle combined with the first term of the second Cycle. This new
60-element Cycle also gets the name niowanggiyan singgeri and all names
are repeated with the period 60.
gorel-1.p65
211
4/17/02, 8:57 AM
212
part four
The 60-element Cycle is represented by the following terms: 1. 2. 3. 4. 5. 6. 7. 8. 9. 10. 11. 12. 13. 14. 15. 16. 17. 18. 19. 20. 21. 22. 23. 24. 25. 26. 27. 28. 29. 30.
niowanggiyan singgeri niohon ihan fulgiyan tasha fulahån gålmahån suwayan muduri sohon meihe àanyan morin àahån honin sahaliyan bonio sahahån coko niowanggiyan indahån niohon ulgiyan fulgiyan singgeri fulahån ihan suwayan tasha sohon gålmahån àanyan muduri àahån meihe sahaliyan morin sahahån honin niowanggiyan bonio niohon coko fulgiyan indahån fulahån ulgiyan suwayan singgeri sohon ihan àanyan tasha sahahån gålmahån sahaliyan muduri sahaliyan meihe
31. 32. 33. 34. 35. 36. 37. 38. 39. 40. 41. 42. 43. 44. 45. 46. 47. 48. 49. 50. 51. 52. 53. 54. 55. 56. 57. 58. 59. 60.
niowanggiyan morin niohon honin fulgiyan bonio fulahån coko suwayan indahån sohon ulgiyan àanyan singgeri àahån ihan sahaliyan tasha sahahån gålmahån niowanggiyan muduri niohon meihe fulgiyan morin fulahån honin suwayan bonio sohon coko àanyan indahån àahån ulgiyan sahaliyan singgeri sahahån ihan niowanggiyan tasha niohon gålmahån fulgiyan muduri fulahån meihe suwayan morin sohon honin àanyan bonio àahån coko sahaliyan indahån sahahån ulgiyan
3.10. Words for Periods of Time In Manchu, there are particular words for certain days of the month and for months and years. Months are correlated with the cardinal numerals, with the exception of three months, which have special names: aniya biya (also tob biya) “the first lunar month;” omàon biya “the eleventh month;”
gorel-1.p65
212
4/17/02, 8:57 AM
213
morphology
jorgon biya “the twelfth month” (Zakharov, 1879:102). The first month of one of the four seasons is indicated by the word uju “head,” “first” (NL:293). In addition, each month is correlated with a certain term of the 12-element Cycle: 1) 2) 3) 4) 5) 6) 7) 8) 9) 10) 11) 12)
the first month the second month the third month the fourth month the fifth month the sixth month the seventh month the eighth month the ninth month the tenth month the eleventh month the twelfth month
tasha biya gålmahån biya muduri biya meihe biya morin biya honin biya bonio biya coko biya indahån biya ulgiyan biya singgeri biya ihan biya
A month is divided into three ten-day periods, and each of these periods has its own name. The name for the first period is ice “new,” and a date is formed by this word and one of the cardinal numbers from one to ten following it: ice juwe “(to-day) is the second,” ice juwan “(to-day) is the tenth.” The second and third ten-day periods are called juwan deri and orin deri (correspondingly). Some days of the month also have their own names: ice “the first day of the month;” àongge inenggi “the first day of the lunar month;” aniya inenggi “New Year’s day;” ilangga inenggi “the third day of the third moon;” jorgon inenggi “the eighth day of the twelfth month;” sunjangga inenggi “the fifth day of the fifth month;” nadangga inenggi “the seventh day of the seventh moon;” wangga inenggi “the fifteenth day of a lunar month;” niolhun “the sixteenth day of the first moon,” “the end of the new year’s festivities;” sucungga nadan “the first seven days after death” (Zakharov, 1879:102; Möllendorff, 1892:7; NL). Some years also have their own names, for instance: sucungga aniya “the first year (of the emperor’s reign).” Hours (ma. erin) of the twenty-four hour period are named after the twelve animals (see the12-element Cycle). An hour is divided into kemu “a quarter of an hour,” fuwen “a minute” and miori (miyori) “a second.”
gorel-1.p65
213
4/17/02, 8:57 AM
214
part four 4. Pronouns
The term “pronoun” (pronominal word) is traditionally used to de scribe a wide range of words united by a common deitic function or its variety, i.e. the function of substitution (anaphoric function). In contrast to nouns, the pronominal words do not name persons, objects or attributes, merely refer to them. For example, the word “he” refers to a male, the words “what,” “which,” “such” may re fer to any attributes, and the words “there,” “here,” “when,” “how” are used to substitute adverbs denoting place, time or manner of per forming an action (anaphoric function). In a certain sense, the deicticsubstitutional words, that is traditional pronouns and pronominal adverbs, form a special system which is parallel to that of the nomi native parts of speech and which, in miniature, duplicates the lat ter. Even substitutes for verbs, so-called pronominal verbs, for ex ample, exist: the verb “to do” in English, the verb “lai” (literally: “to come”) in Chinese, etc. (Maslov, 1987:166-7). There are two possible ways of classifying all pronominal words. Pronominal words can be divided into noun-pronouns, pronominal adjectives, indefinite numerals (pronominal numerals), and pronomi nal adverbs according to the parts of speech they belong to. Another way to classify pronominal words is to distribute them among dif ferent semantic groups on the grounds of their semantic affiliation. It is accepted by most specialists that all Manchu pronominal words can be divided into several semantic subgroups. These are personal, possessive, demonstrative, interrogative, and indefinite pronouns. Various scholars include words that differ greatly from the seman tic point of view, e.g. both words that have quantitative and quali tative semantics, in the indefinite pronouns class and hence this subgroup of pronouns is the most problematic. The word beye “body,” “self” in different forms is used to express reflexive meaning.
4.1. Personal Pronouns. Irregular Pronominal Case Forms The personal pronouns are bi “I,” “me;” si “you (singular);” i “he,” “she;” be “we (exclusive);” muse “we (inclusive);” su(w)e “you (plural);” ce “they.” Not only the first and the second but also the third per son pronouns are used to refer to human beings. These pronouns cannot be used to refer to objects and qualities. To refer to objects
gorel-1.p65
214
4/17/02, 8:57 AM
215
morphology
and qualities the demonstrative pronouns ere “this” and tere “that” are normally used. In Manchu, as in other Tungusic languages, there are two forms for the first person plural, i.e. be and muse. Be “we” (1 person plural exclusive) is used to refer to the number of referents, 1 ps. + 3 ps., from which 2 ps. (the interlocutor or the person spoken to) is ex cluded. Muse “we” (1 person plural inclusive) is used to refer to the number of referents, 1 ps. + 2 ps., i.e. consisting of the speaker and the listener. According to Möllendorff, muse means “we that are speaking together, we that belong to one family, one clan, one nation” (Möllendorff, 1892:6). In his Grammar, Zakharov cited a very typical sentence, taken from the Russian-Chinese Treaty of 1728: ne men-i hinggun tugurik gebungge now we(EXCL)-GEN=our place-name place-name named ba-de suwen-i niyalma kemuni jecen place-DAT you(PL)-GEN=your people constantly border daba-me yabu-mbi, ere-be muse gurun cross-CONV go-IMPF this-ACC we(INCL) country lashala-me gisure-rakå o-ci, make.a.decision-CONV speak-PART(NEG) be(AUX)-CONV amala ere jergi jecen i ba-de te-he urse later this other border GEN place-DAT live-PART people ishunde urunakå baita dekdebu-re de to.one.another necessarily matter rise.again-PART DAT isina-mbi go.as.far.as.to-IMPF “Now at our place, named Hinggun Tugurik, your people constantly cross the border. If we (INCLUSIVE) two countries do not make a decision, then later people who live at this and other frontier places will be forced to take up the quarrel again” (ZAKH:113). Personal pronouns are declined similarly to nouns, but there some peculiarities in their declension. The personal pronouns for the first, second and third persons, singular and plural, with the exception of muse “we (inclusive),” have some irregular case forms, i.e. min-, sin-, in-, men-, su(w)en-, cen-. The stems of the personal pronouns in the oblique cases, unlike in the nominative case, terminate in the
gorel-1.p65
215
4/17/02, 8:57 AM
216
part four
consonant n. Alternation of the consonants b and m in the forms for the first and second persons, viz. bi/min- and be/men-, could be explained by regressive assimilation, due to which the initial b un der the influence of the final n has changed into m. It is important to mention that all Tungus-Manchu languages have irregular case forms. In classical Manchu, irregular pronominal case forms ending in -n change their final consonant into -m under the influence of the initial consonant of the marker for the accusative which is be: mimbe < min- “I,” “me” + be; simbe < sin- “you (SG)” + be; suwembe < suwen“you (PL)” + be; imbe < in- “he,” “she” + be; cembe < cen- “they” + be. The marker for the accusative is usually written separately from the word it follows, with the exception of the personal, possessive, demonstrative, and interrogative pronouns, as well as monosyllabic words: aibe (ai “what?,” “which?”), babe (ba “place”); erebe (ere “this”), webe (we “who?”). Personal pronouns are declined as shown in table 7. Table 7. Pronominal Case Forms Singular Nominative Genitive Dative Accusative Ablative
bi “I” mini minde mimbe minci
si “you” sini sinde simbe sinci
i “he,” “she” ini inde imbe inci
Plural Nominative Genitive Dative Accusative Ablative
be “we (EXCL)” muse “we (INCL)” meni musei mende musede membe musebe menci museci
su(w)e “you” su(w)eni su(w)ende su(w)embe su(w)enci
ce “they” ceni cende cembe cenci
The meaning of the personal pronouns in the genitive case corre
sponds to that of the possessive pronouns:
min-i ama eme emu mejige bene-fi …
I(bi/min-)-GEN=my father mother one news send-CONV
“(Let you) pass a news to my father and mother … ” (NSB:113);
min-i eigen biài-rakå be
I(bi/min-)-GEN=my husband be-PART(NEG) ACC
gorel-1.p65
216
4/17/02, 8:57 AM
217
morphology sa-rkå know-PART(NEG)
“I do not know whether my husband is absent;”
muse-i horgon de emu morin bi we(INCL)-GEN= our stable DAT one horse COP “There is a horse in our stable” (SK:54). 4.2. Demonstrative Pronouns
The demonstrative pronouns are ere “this” and tere “that.” Plurals are formed by the suffix -se which is attached to the clipped stems e- and te- correspondingly: ese (erse) “these” and tese “those.” The demonstrative pronouns ere and tere have two variants of stems in the dative case. They are declined as shown in table 8.
Table 8. Demonstrative Case Forms Singular Nominative Genitive Dative Accusative Ablative
ere “this” erei, ereni ede, erede erebe ereci
tere “that” terei tede, terede terebe tereci
Plural ese “these” esei esede esebe eseci
tese “those” tesei tesede tesebe teseci
The demonstrative pronouns ere “this” and tere “that” are often used instead of the third person i (singular) “he/she” and ce (plural) “they.” The demonstrative pronouns may substitute all nouns including human ones. In these cases, they are followed by case markers: ere be ai arga i han deri tucibu-me gai-mbi she ACC what trick GEN khan SEP take.out-CONV take.awayIMPF “(We) shall take her away from the khan by some trick” (SK:55); ere be tuwa-ci ... this ACC look-CONV “Looking at this ... ”(ZAKH:114). If serving as attributes, they are not followed by case markers:
gorel-1.p65
217
4/17/02, 8:57 AM
218
part four
tere niyalma beye sabu-ha that man self see-PART “I saw that man myself” (ZAKH:114); enengi (ma. enenggi) ere ba de ainu ebu-mbi today this place DAT why stop-IMPF “Why have (you) stopped today at this place (here)?” (SK:55). In the following sentence the dative form of the pronoun ere “this,” which can be seen as a pronominal adverb, serves as the adverbial modifier of place: e-de ume ebu-re this-DAT NEG stay-PART “Do not stay here” (SK:55). The following two pairs of demonstrative pronominal adverbs may be assigned to this group of pronominal words: uttu “like this”—tuttu “like that” and enteke “like this,” “this kind of,” “such (a)”—tenteke “like that,” “that kind of,” “such (a).” Both pairs of pronominal words, having the qualitative-demonstrative meaning, are used as attributes. Sometimes the pair of pronominal adverbs uba (< *u “this” + ba “place”) “this (place)”—tuba (< *tu “that” + ba “place”) “that (place)” is used in the meaning “this” and “that,” respectively. They are never used in combination with nouns, but always function as nominal substitutes. The pair of indefinite numerals (pronominal numerals) utala “so many (as this)”—tutala “so many (as that)” have quantitative-demonstrative meaning. It is interesting to note that the first components of the pairs “ere—tere,” “uttu—tuttu,” “enteke—tenteke,” “uba—tuba,” “utala—tutala” indicate something that is situated at a short distance from the person who is speaking. They are opposed to the second components which show something that is remote in time and space. Thus, the initial vowels e and u indicate proximity, while the initial consonant t is used to indicate remoteness in space and time (Pashkov, 1963:31). It should be mentioned here that there is evidence that in differ ent versions of the Sibe dialect the demonstrative pronouns tere “that” and tese “those” may substitute the singular and plural forms of the third person i “he,” “she” and ce “they” (Norman, 1974:167; Lebedeva & Gorelova, 1994:54).
gorel-1.p65
218
4/17/02, 8:57 AM
219
morphology 4.3. Possessive Forms of the Personal and Demonstrative Pronouns
The possessive meaning is expressed by adding the genitive to the personal pronouns which denote the first, second and third persons, singular and plural: mini boo “my house,” meni (musei) boo “our house,” sini boo “your (SG) house,” su(w)eni boo “your (PL) house,” ini boo “his/ her house,” ceni boo “their house.” If used alone, not in combination with nouns, these pronouns may be followed by the possessive suffix -ngge or -ningge (which is used here as a substantivizer): miningge “mine,” siningge “your (SG),” etc. In this case, they are declined similarly to nouns. They are mostly used in anwers when the object which is in possession, is already known from the question: ere uthai we-i jaka this then who-GEN=whose thing “Whose is this thing?” miningge mine (ZAKH:114). Demonstrative pronouns in the form of the genitive also take the possessive suffix -ngge: ereingge, eseingge, tereingge, teseingge. Like the personal pronouns followeded by the possessive suffix, they are never followed by nouns. As far as uttu and tuttu are concerned, they are followed by the possessive suffix in the form of the nominative (di rect) case: uttungge, tuttungge. 4.4. Interrogative Pronouns The interrogative pronouns are represented by words we “who?” (gen. wei “whose?;” dat. wede “to whom?,” “for whom?;” acc. webe “whom?;” abl. weci “from whom?”); ai “what?,” “which?” (gen. aini “wherewith?,” “whereby?;” dat. aide “where?,” “whither?,” “why?,” “how?;” acc. aibe “what?;” aici “what sort of ... ?”); ya “what?,” “which?” (dat. ya de “where?,” “whither?;” acc. yabe; abl. yaci “from where?”); eke “who is that?;” udu “how many?,” “how much?”. These basic interrogative pronouns are used to form more com plex pronouns, such as: aba “where?;” aiba (< ai + ba “place”) “where?” (dat. aibade “where?;” abl. aibaci “whence?,” “from where?”); aibi (< ai + bi=copula) “where?,” “what is there?” (dat. aibide “where?,” abl. aibici “whence?,” “from where?,” sep. aibideri “from
gorel-1.p65
219
4/17/02, 8:57 AM
220
part four
where?”); ainu (< ai + nu = interrogative particle) “how?,” “why?;” the pronominal words aika (< ai + eke), aimaka (< ai + maka), yamaka (< ya + maka) (maka is explained by Norman as “an introductory particle of doubt or questioning: I wonder ... , could it really be that ... ;” NL:192) may have the interrogative meaning as well the indefinite meaning of “any,” “some” (see 4.5); ekehe (< eke + bihe = the perfect participle of the verb bi- “to be”) “who was that?;” weingge (< we + i = the genitive + -ngge = the possessive suffix) “whose?;” weke (< we + eke) “who is he?;” yaba (< ya + ba “place”) “where?;” yaka (< ya + eke) “someone, who?”. The words aimaka and yamaka may also display the adverbial meaning “probably,” “seemingly.” The interrogative meaning may be also expressed by the pronomi nal adverbs adarame “how?,” “why?,” “how so?” and atanggi “when?”. There are several patterns in which some pronouns combine with each other: we ai, we ya “who?”. These pronouns are used with re gard to a number of persons, however, the semantic difference between them is not clear. It seems that another pair of pronouns eke ya “who was it now?” is used when someone cannot think of a person’s name (NL:72). When the interrogative pronouns are declined, they are commonly written together with the case marker as one word: aibe (< ai + be = accusative) “what?;” aici (< ai + ci = ablative) “what sort of?;” aide (< ai + de = dative) “where?,” whither?,” “why?,” “how?;” webe (< we + be = accusative) “whom?,” wede (< we + de = dative) “to whom?,” “for whom?;” yaci (< ya + ci = the ablative marker) “from where?”. Occasionally, they are written separately, for example: ya de “ where?,” “whither?”. The most commonly used interrogative pronouns are ai and ya. They may serve as the question to all nouns without exception. The abovementioned meanings are attributed to them only in context, when combined with words belonging to different semantic groups. There are certain differences in their meanings. The pronoun ai “what,” “how” is used when an inquiry is made about the quality of an object: ai jugån “what road” (of high or poor quality?). When combined with a noun that denotes a place, the pronoun ya “what?,” “which?” usually has the meaning of “which?”. The pronoun ya in the dative and ablative cases serves as a question to words which function as adverbial modifiers of place: ya de “where?,” yaci “from where?”. Followed by the same case markers, the pronoun ai serves as a question to words which denote indirect objects: aide “in what?,”
gorel-1.p65
220
4/17/02, 8:57 AM
221
morphology
“to whom?;” aici “from whom?,” “from what?”. These differences in meaning are often subtle and difficult to discern. Here are some
examples with the pronoun ai: ài (ma. si) ai nimeku baha-fi you what illness get-CONV “What illness did you get?;” bira-i cargi de ài (ma. si) ai be sabu-ha river-GEN there DAT you(SG) what ACC see-PART “What did you see over the river?” (SK:55). The following examples contain the pronoun ya: ya ici gene-mbi, ja jugåm-be jabu-mbi, jetere jefelin which side go-IMPF which road( jugån)-ACC go-IMPF food food ya gese baitala-mbi ai jaka gama-mbi which like use-IMPF which thing take-IMPF “Which side shall (I) go towards , which road shall (I) go along, which food shall (I) use, which things shall (I) take?” (SK:55-6). Sometimes the pronouns ai and ya are used in relative function: ya ici genere be sa-rkå which side go-PART ACC know-PART(NEG)
“(Someone) does not know towards which side he will go;”
ama àin-de (ma. sin-de) ai se-ci ai-be bu-ki
father you(si/sin- SG)-DAT what mean-CONV what-ACC give-OPT
“The father will give you (everything) what you want” (SK:56).
The pronoun we “who?” is exclusively used to refer to human be
ings. It can have both a deictic and an anaphoric functions. The
pronoun wei “whose” is also used to refer to human beings. Since
only human nouns, whose referents are human beings, have mor
phological plurals the interrogative pronoun we also has plural formed
by compounding pronominal words: we ai and we ya “who (plural).”
Like ai and ya, the interrogative pronoun we is used in relative func
tion:
gorel-1.p65
221
4/17/02, 8:57 AM
222
part four
ere morim-be we jafa-me mute-ci, we de this horse(morin)-ACC who catch-CONV can-CONV who DAT min-i sain sargan jui de bu-mbi I(bi/min-)-GEN = my beautiful woman daughter DAT give-IMPF “I shall give my beautiful daughter to whoever can catch this horse” (SK:57). The interrogative pronoun ainu “how?,” “why?” is derived from the pronoun ai and the interrogative particle nu. Here are some instances of the use of the interrogative pronoun ainu: ainu uttu ere ba-de te-hebi why so this place-DAT sit-PAST “Why did (you) sit so (in such a manner) at this place?;” ài (ma. si) ainu amala o-ho you why behind be-PART “Why have you been behind (everyone)?” (SK:57). The interrogative pronoun udu “how much?,” “how many?” serves to indicate quantitative characteristics of objects in the form of a question: udu niyalma “how many persons?” and udu inenggi “how many days?”. Here is an examples: ài (ma. si) ere udu inengi (ma. inenggi ) ya de you this how.many day which DAT jabu-ha walk-PART “Where did you walk for so many days?” (SK:57). The verbs aina- “to do what?” and aise- “to say what?,” “to be called what?” are derived from the interrogative pronoun ai. The form of the perfect participle of the verb aina- “to do what?” which is ainaha, shows the tendency to shift from participles to interrogative pronouns: ainaha “what happend?,” “what sort of?,” “what kind of?”. It mostly occurs in interrogative sentences such as: ere ainaha mahala? “What sort of hat is that?;” ài ainaha niyalma? “What sort of man are you?” (SK:57).
gorel-1.p65
222
4/17/02, 8:57 AM
223
morphology 4.5. Indefinite Pronouns
The group of indefinite pronouns has developed historically to in clude several subgroups of pronominal words that may be differen tiated by their semantics. Some of them actually have the indefinite meaning: aika “any,” “something;” aimaka, yamaka “any,” “some;” ememu “some;” ememungge “some” (as substantive); pronominal adverb ududu “several,” “a number of,” motivited by the interrogative pro nominal adverb udu “how much,” “how many.” Others have a col lective meaning: ganji “all” (this word may also function as an ad verb with the meaning “completely”); gemu “all” (this word may function as an adverb with the meaning “even”); geren “all,” “many;” gubci “all,” “entire;” ele “all,” “all, who,” “all, which” (used after par ticiples); yooni “all,” “all together,” “complete,” “entire.” There are others that denote many homogeneous units, such as: eiten “all” (in attributive position), “every;” yaya “every,” “each.” The words dari and tome also have the meaning “every,” “each,” however, they can more rightfully be classified as postpositions rather than pronouns since they normally follow nouns, and precede the governing verbs. Some pronominal words, with the same meaning, are formed by repeating the word, that is by reduplication: beri beri “each one;” meni meni “each,” “every;” teisu teisu “one by one” (this reduplicated word may display the meaning “all together,” “on every occasion”). The pronoun yaya is also formed from the reduplicated pronoun ya ya, the components of which had lost their phonetic independence. Pronominal words that have demonstrative meaning of “not this,” “not that,” “not like this (that)” have also been traditionally included in this group of pronominal words: gåwa “other,” “another;” gåwaingge “someone’s else;” weri “another,” “other,” “somebody else.” According to Lebedeva, the pronoun “meni meni” preserved the independence of components from which it is formed, but it had lost connections with the word from which it is derived. In classical Manchu, the word meni “our” is found, however, it is a homonym of the word meni which is appears in the reduplicated pronoun “meni meni.” The latter’s component meni goes back to the pronoun mon “self,” “one’s” which is widely used in the languages of the Nanai group. This component meni does not occur either in classical Man chu or in Sibe, therefore it seems to be borrowed from some dialect or cognate language. Semantically it is very close to the word beye “body,” “self,” when the latter is used as pronoun (Lebedeva &
gorel-1.p65
223
4/17/02, 8:57 AM
224
part four
Gorelova, 1994:58). The following sentence contains the pronoun meni meni: men’i men’i tere dedu-re bou (ma. boo) de bi-bu-he that bed-PART house DAT stay-CAUS-PART meni meni “each,” “every,” (NL:197); meni meni “every one,” each one,” “individual” (YAM:145); men’i men’i “every one,” “individual” (SK:170); dedure boo “bedroom;” “(They) placed everyone in bedrooms” (SK:58). When these pronouns serve as attributives, they are not followed by case markers. Used as substantives, they are followed by case formants: geren be ji-bu-kini se-he all ACC come-CAUS-OPT say-PART “(He) commanded that everyone should come (there)” (ZAKH:116). According to Zakharov, the words dari, gemu, tome normally appear after the case marker following the preceding noun and in turn followed by the main verb. Here are some examples: niyalma de gemu bi person DAT all have “All people have (smth.)” (ZAKH:116); ere hotun (ma. hoton) gupci (ma. gubci) ajige amba niyalma be this walled city all small big people ACC gemu soli-fi all invite-CONV “All people, small and big, were invited to this city” (SK:58). The expression ajige amba niyalma “small and big people” is framed by two semantically identical words, the first of which should be analyzed as an indefinite pronoun and the second one as a postposition. 4.6. The Relative Meaning of the Indefinite Pronoun “ele” The indefinite pronoun ele may have a relative meaning when it oc curs between two words, the first of which is used as an attributive
gorel-1.p65
224
4/17/02, 8:57 AM
225
morphology
and the second as a substantive. In the case when the substantive is absent, the word ba “place,” “circumstances,” “occasion” functions as a substantivizer. The pronoun ele is frequently used after the imperfect and perfect participle forms, which function as attributives. Here are some examples: min-i ala-ha ele ba I(bi/min-)-GEN tell-PART all SBSTR “All that I told;” gånin adali ele niyalma thought same all person “Everyone who (has) the same thoughts;” ji-he ele niyalma come-PART all people “All the people who came” (ZAKH:117). The pronoun ele and the participle are often written as one word. The initial e of the pronoun ele is dropped, and the participle is followed by the syllable le, the vowel of which is changed in accor dance with the law of vowel harmony: isina-ha-la ba reach-PART-REL place “All places which (smb.) reached;” bisi-re-le baita be-PART-REL matter “All matters which exist;” donji-ha-la urse urgunje-rakå-ngge akå hear-PART-REL people joy-PART(NEG)-NR there.are.not “There is no one who has head (smth.) and does not rejoice” (ZAKH:117). In old Manchu the syllable le usually does not follow the law of vowel harmony: donjihale, isinahale (Zakharov, 1879:117). 4.7. Pronominal Use of the Word “ beye” In classical Manchu there are no reflexive or reflexive-possessive pro nouns. The word beye “body,” “self” is used instead. Being a noun, this word is declined in the same way as declinable nouns. The plural
gorel-1.p65
225
4/17/02, 8:57 AM
226
part four
is formed by the suffix -sa (-se): beye “oneself”—beyesa “themselves.” The declension of the word beye is shown in table 9. Table 9. Declension of the word beye
Nominative Genitive Dative Accusative Ablative
Singular
Plural
beye beye(i) beye de (beyede) beye be (beyebe) beye ci (beyeci)
beyesa beyese beyese de (beyesede) beyese be (beyesebe) beyese ci (beyeseci)
When functioning as the reflexive pronoun the word beye is charac terized by the following features: 1) the word beye follows the personal pronouns which are used in the genitive case: Singular mini beye “myself” sini beye “yourself” ini beye “herself”/”himself”
Plural meni beye (beyese) “ourselves” musei beye (beyese) “ourselves” su(w)eni beye (beyese) “yourselves” ceni beye (beyese) “themselves”
2) in the plural there is grammatical agreement in number: instead of mini beyese the combination meni beyese is used. In the nominative case, used before a verb, the word beye has the reflexive meaning “oneself:” bi tere niyalma be beye sabu-ha I that man ACC myself see-PART “I saw that man myself” (ZAKH:115); ere han-de bi beye aca-me gene-mbi that khan-DAT I myself meet-CONV go-IMPF “I shall go to meet this khan myself” (SK:58). In the genitive case the word beye may be used as the reflexive-possessive pronoun “one’s own:”
gorel-1.p65
226
4/17/02, 8:57 AM
227
morphology beyei boo de dosi-ka his house DAT enter-PART “(He) entered his own house” (PASH:31).
Followed by the originally possesive suffixes -ngge or -ningge (which functions here as a substantivizer), the word beye displays the reflexive-possessive meaning of “one’s own” as a substantive: – ere uthai wei jaka?this namely whose thing – “This thing is whose?”– miningge (beyeiningge)– “Mine (own)” (PASH:31). Followed by a case marker, this word in its reduplicated form, beye beye, has the meaning “between oneselves,” “each other,” “one another:” ninggun niyalma beye beye i ici gisure-hei six man each other GEN with speak-CONV “Six men spoke with each other for a long time.” 4.8. Nouns in the Pronominal Function In Manchu, as in other Oriental languages, the use of personal pro nouns, especially for the first and second persons, was avoid by people in educated circles. Instead, following the Manchu-Chinese etiquette, the Manchus used combinations of personal pronouns in the genitive case with the word beye: mini beye “my person,” sini beye “your per son.” In addition to these native devices, the Manchus borrowed sub stitutes for personal pronouns from Chinese. Following the Chinese language, the Manchus used two groups of words which could serve as substitutes for the first and second persons. The words of the first group concerned a person who addressed an interlocutor. These particular words indicated politeness, humility, and submission and can be referred to as self-deprecatory terms. By contrast, with re gard to the second person, the special words of entitling as well as those which denoted posts, rank, and status, viz. polite terms of address, were used (Zakharov, 1879:108-12). The self-deprecatory terms were used in the following situations.
gorel-1.p65
227
4/17/02, 8:57 AM
228
part four
Addressing subjects, the Emperor, referring to himself, used the pro noun bi “I,” “me” and mini beye “myself.” Addressing princes (in Manchu: amban and beile), the Emperor used such terms as sitahån niyalma “poor man” (sitahån “deficient,” “scarce”) and emteli beye “orphan” (emteli “alone,” “sole”). Addressing the Emperor, his princes, referring to themselves, used the following terms: sitahån niyalma “poor man,” emteli beye “orphan,” and emhun beye “orphan,” “lonely person” (emhun “alone,” “sole,” “lonely”). The word amban in the sense of “servant of your majesty” was used as well. Addressing the ruler, the most close descendants of the ruling Manchu dynasty used the word aha “slave.” Accord ing to the same pattern, addressing the ruler or high officials the Chinese originally used the word amban, but the Manchus used the word aha. Addressing subordinates, high officials used the following terms: adali hafan “equal official,” emu ba i hafan “fellow-clerk” (literally: the same place’s official), uhei deo bi “I am a common younger brother” (uhe “common,” “mutual,” “of one kind” + i = GEN). Addressing officials of higher rank, those of lower rank used the following turns of speech: buya tuàan i hafan “official of seventh or eighth ranks” (literally: insignificant post’s official );” hartu or harangga hafan “subordinate official.” Addressing the ruler, a common person should use such a selfdeprecatory term as buya irgen “small person,” “ordinary person.” Addressing officials of any rank, a common person should use the following self-deprecatory terms: irgen niyalma “common person,” fusihån beye “subordinate person,” buya beye “insignificant person.” Persons, who were equal in age, status or rank, as well as those who were friends, when addressing each other, instead of bi “I” used the words deo “younger brother” and buya deo “insignificant younger brother.” Instead of the pronominal substitute for the person to whom the speech was addressed (addressee or interlocutor), viz. si “you (SG),” they used such polite terms of address as ahån “elder brother,” age, agu = respectful term of address for men: master, sir, lord (< turk. aga “elder brother”). In case when the person who was speaking could not refer to himself by self-deprecatory terms due to old age or social status, he would normally use the term mentuhun ahån “stupid or silly brother.” The word mentuhun “silly,” “stupid” was often used by younger rela tives when they addressed the elders, for example: mentuhun jalahi jui
gorel-1.p65
228
4/17/02, 8:57 AM
229
morphology
“silly nephew (son of one’s brother),” mentuhun omolo “silly grand son.” Persons, equal in status, without indicating age or kinship, mostly used the following self-deprecatory terms: fusihån beye “subordinate person,” ajige beye “small person,” mentuhun bi “I am a silly person,” albatu niyalma “common, ordinary person,” sesheri beye “vulgar, com mon person.” The words of the second group, being substitutes for the second person, were used to entitle or honour the interlocutor. These hon orific terms were used in the following situations. The entitling of the Emperor followed a strict Manchu-Chinese etiquette. Addressing the Emperor, such native Manchu words as han “emperor,” “khan” and ejen “emperor,” “ruler,” “lord,” “mas ter” were used. In addition to this, there were loan-translations borrowed from Chinese: abkai jui “the Son of Heaven,” “the Em peror.” Two terms dergi “the Emperor” (literally: “upper”)” and dele “the Emperor (literally: top)” corresponded with the term “your Majesty.” The calques tumen se enduringge ejen “ten thousand years— long living- the divine lord,” “the Emperor;” genggen ejen “an enlight ened ruler” were used as well. The Chinese term for the Emperor håwangdi (< huang di) was widely used by the Manchus8. When addressing the successor to the throne (the Crown Prince), the Chinese term håwang taidz “Heir Apparent” was used. Other princes were called also by the Chinese term taidz “Heir Apparent (literally: high son).” Very young princes were called by the Manchu word age “Prince, Son of an Emperor” (also a polite term of address: “master, sir, lord”). When referring to female persons who belonged to the ruling Manchu Dynasty, the Chinese terms håwangheo “empress,” gungju “princess,” håwang guifei “an imperial concubine of the first rank,” fei “an imperial concubine of the third rank” and pin “an imperial concubine of the fourth rank” were used. When referring to princes and others who possessed hereditary titles, the Chinese terms wang “prince,” gung “duke” and the Manchu words beile “ruler,” “prince of the third rank” and beise “ruler,” 8 Subordinated to China, foreign rulers such as the Mongolian princes and the Korean king were addressed by the word han “khan.” However, this word had a different spelling from the one used to address the Manchu emperor. By con trast, the final n in the title for the Manchu emperor was written with a dot (Zakharov, 1879:110).
gorel-1.p65
229
4/17/02, 8:57 AM
230
part four
“prince of the fourth rank” were used. These terms were combined with such words as tob “straight,” “right,” “upright;” mergen “wise” or beye “person,” for instance: wang ni beye “person of high rank.” When addressing civil and military officials of the first four ranks, the word amban “high official,” “dignitary” or the combination of amban and wesihun “honorable,” “respected,” viz. wesihun amban, were used. When addressing officials of low rank, people used the Chinese word laoye “master,” “bureaucrat,” “overlord” preceding this term by the Manchu word amba “high:” amba laoye. When addressing persons equal in age or social status, the fol lowing terms with the meaning “you (honorific)” were used: wesihun beye, wesihun nofi “respected (high) person” or agu, age which was used as a polite or respectful term of address, and ahån i beye “elder brother.” When addressing someone in letters, people used such polite or respectful terms as ambalinggå beye “dignified person,” yekengge beye “noble, grand person.” When addressing a serviceman, the follow ing honorific terms were used: baturu haha “brave man,” “hero;” aisin cira “dear person” (literally: golden face);” wesihun cira “respected person (literally: respected face);” fujurungga cira “noble person (lit erally: noble face).” The most used of the honorific terms were those relating to seniority of brotherhood: gosingga ahån (literally: cherished elder brother), amba ahån (literally: big elder brother), ahångga ahun (literally: eldest elder brother). When referring to teachers, medical doctors or artisans of vari ous kinds, the word sefu derived trom the Chinese word shifu “mas ter in trade, business or any troupe who undertakes to teach skill to pupils” (CD:397) was also used as a polite form of address to people who had skills or specialized knowledge. The expression sagda sefu “an old master” was also used. When referring to prince’s wives and other female persons who possessed hereditary titles, for instance, wife of a feudal lord (ma. beile), the word fujin derived from the Chinese word furen “Mrs,” “madam,” “lady” (CD:132) was used. The respectful term of address for young ladies was the Manchu word gege “elder sister.” The re spectful terms of address for wives of officials of various kinds were such words as the above-mentioned fujin and the word taitai again borrowed from the Chinese word taitai “Mrs,” “madam” (CD:430). Native Manchu words were also used: eniye “mother,” nainai and niyangniyang “housewife,” “mistress of a household.” According to
gorel-1.p65
230
4/17/02, 8:57 AM
231
morphology
Norman, the word niyang means “girl,” and the reduplicated form of this word niyang niyang (niyangniyang) must be translated as “godess” (NL:217). The respectful term for young ladies jiye jiye was borrowed from Chinese. It is the reduplicated form of the Chinese word jie “elder sister” (CD:228) which is used as a general term of address for young women. Norman gives the lexical entry jiyei with the trans lation “older sister,” “miss” (NL:161). In official relations between State ministries, boards, and depart ments and officials who worked there, expressions derived from the word wesihun “upward,” “honorable,” “respected” and words denot ing government offices, posts, positions, and appointments were used. Here are some examples: wesihun jurgan “respected ministry;” wesihun yamun “respected office;” wesihun ba “respected department (place);” wesihun amban “respected dignitary;” wesihun hafan “respected officer.” The afore-cited style of using substitutes for the first and second persons was the characterizing feature of the spoken language of the nobility and the educated circles. Townspeople tried to imitate this style of speaking in order to seem more refined and sophisticated. It was not the custom of the common people to follow this style of speech and they consequently used all personal pronouns without exception (Zakharov, 1879:112).
5. Verbal Morphology In the Manchu grammatical structure the verb plays a significant role in the formation of an utterance. Three types of verbal forms, derived from the same stems by the adding of particular suffixes, exist in Manchu. Therefore each type is characterized by its own paradigm of suffixes. These forms are participles, converbs, and verbs proper. No verbal forms have morphological categories of gender or even markers denoting bio logical sex of living beings, person or number. This means that the verbal forms cannot be followed by suffixes which denote these grammatical categories. These notions are conveyed by the seman tic and syntactic context with the help of grammatical characteris tics inherent in nouns and pronouns. Thus, the verbal forms deter mine their relationship with the category of sex through those characteristics which are used in the language to denote this mean ing by nouns. The category of number, singular and plural, is co
gorel-1.p65
231
4/17/02, 8:57 AM
232
part four
ordinated with verbal forms through that of nouns and pronouns. The category of person is made evident in verbal forms by the use of personal pronouns, singular and plural, as well as other pronouns which denote plurality. A person can also be determined by con text, especially in short replies to preceding questions. According to their attitude to the category of direct object, all verbs can be subdivided into transitive or intransitive. In Manchu, we should distinguish between the predicative copula bi which is likely of non-verbal nature and the existential verb bi“to exist,” “to be.” The copula bi, which has affirmative modal mean ing, together with the other copulae (they are described in Section 9, specially devoted to them), plays an important role in the Manchu sentence. In certain cases, it is an obligatory structural component of the predicate, especially in the sphere of the nominal type of predication. The verb bi- changes in accordance with the grammatical categories of time and mood. A number of non-finite verbal forms can be derived from this verb. The most frequently used ones are: the indicative forms bi-mbi, bi-mbihe, bihebi; the imperative forms bisu, bi-kini, bi-cina, bisi-reo; the optative form bi-ki; the participial forms bi-si-re, bi-si-re-ngge, bi-si-rakå, bi-si-rakå-ngge, bi-he, bi-he-ngge, bi-hekå, bi-hekå-ngge; and the converbal forms bi-me, bi-fi, bi-hei, bi-ci, bi-cibe, etc. It is important to mention, that normally the copula bi may be combined with one of the forms of the verb bi-. Moreover, frequent use of the copula bi with certain non-finite verbal forms may lead to the formation of new verbal forms. Thus, the past form in -habi/ -hebi/-hobi is derived from the form of the perfect participle and the copula bi: -ha/-he/-ho + bi > -habi/-hebi/-nobi. The capacity for amalgamating with various forms of the verb bi- is a specific feature of the copula bi and a productive means for creating new grammatical forms in the language. In Manchu there is a number of verbs which under certain cir cumstances may lose their lexical meaning. Becoming partly desemantisized, they may serve as auxiliary verbs. Apart from the existential verb bi-, the auxiliary function of which is very impor tant, the most frequently used ones are: o- 1) “to become,” “to change into;” 2) “to be,” “to exist;” 3) “to be proper,” “to be permissible,” and se- “to say,” “to call,” “to mean.” Combined with non-finite verbal forms, these auxiliary verbs produce a considerable number of modal and temporal forms of the verb proper, as well as various
gorel-1.p65
232
4/17/02, 8:57 AM
233
morphology
analytical constructions which will be described below, in Section 5.9. 5.1. Verbal Parts of Speech All verbal stems may take participial, converbal or verbal suffixes, and in such morphologically modified forms they function normally in a sentence. The exception is the form of the second person, sin gular and plural, of the Imperative mood. This form coincides with the verbal stem. It may be concluded that this form of the impera tive is marked by zero as it is contrasted with the other moods. The existence of unsuffixed verbal stems is a peculiarity of the Manchu language. No other Tungus-Manchu language is characterized by such a feature. The verb proper and the converb are monofunctional parts of speech, i.e. they have only one function in a sentence. Verbs proper always serve as predicates in simple sentences and in main clauses of complex sentences. Converbs serve as principal predicative mem bers of converbal constructions. Consequently, there is no necessity to use any additional morphological markers to express these syn tactic functions. The grammatical function and semantic relations with other parts of a sentence are expressed by invariable verbal suffixes which are markers of the mood and the converb. The participle is a polyfunctional verbal form which means that participles perform several roles in a sentence. Therefore additional devices are required to express each syntactic function. These par ticular devices are case markers and the nominalizing suffixes all of which will be analyzed below, in Sections devoted to participles (5.6.3 & 5.6.5). 5.2. The Structure of a Verbal Stem and Word-Formation of Verbs A verbal stem consists of a root, derivative suffixes, which belong to the sphere of verbal word formation, and suffixes which denote pe culiar verbal characteristics, i.e. grammatical categories of aspect, voice and mood, including tense, and in some cases indicating a per son. According to their structure the verbal stems may be non-derived and derivative. Most non-derived verbal stems are monosyllabic or disyllabic: bai- “to seek,” “to look for;” bi- “to be,” “to exist;” bu
gorel-1.p65
233
4/17/02, 8:57 AM
234
part four
“to give;” da- “to burn,” “to blow (of the wind),” “to rain,” “to snow;” je- “to eat;” ji- “to come;” o- “to be,” “to become;” sa- “to know,” “to understand;” se- “to say,” “to mean;” te- “to live,” “to sit;” wa“to kill;” ara- “to write;” buce- “to die;” gele- “to fear;” jafa- “to at tack,” “to fight;” mute- “can,” “to be able;” tuwa- “to look;” ucu- “to sing.” Derivative stems are produced in accordance with different word formation patterns. Verbal stems may be derived from nominal words, non-derived verbs, adverbs, and onomatopoeic words by adding semantically diverse derivational suffixes. In Manchu there is a considerable number of stems which serve as the derivative base for nouns and verbs alike. The existence of such stems is a distinc tive characteristic of all the Tungus-Manchu languages. In her de scription of the Tungus (Evenki) language, G.M. Vasilevich referred to such stems “undifferentiated” (Vasilevich, 1940). In fact, in some cases it is doubtful whether the verb is derived from the nominal word or whether the latter is the result of the process of verbal derivation, for example: aca-n “harmony,” “union,” “meeting”—aca“to meet,” “to meet together,” “to combine;” aci-n “load,” “burden”—aci- “to load;” bele-n “false accusation,” “a treacherous murder”—bele- “to harm an innocent person through a false accusation,” “to murder treacherously;” buhiye-n “a guess,” “a surmise”—buhiye“to suspect, “to surmise,” “to guess;” do-n “alighting (of birds)”— do- “to alight (of birds and insects);” fali-n “tying,” “binding”—fali“to tie,” “to bind;” isa-n “assembly,” “a meeting”—isa- “to meet; jila n “compassion,” “pity,” “love”—jila- “to pity,” “to have compassion for,” “to love;” tuksi-n “alarm,” “anxiety”—tuksi- “to be alarmed,” “to be anxious,” “to be afraid,” “to be proud,” “to throb” (Möllendorff, 1892:8). An interesting and distinctive feature of most derivational suffixes is the fact that their semantics permits them to perform not only as derivational suffixes, but also to have other functional meanings. When added to verbal stems, these suffixes denote various aspectual meanings which can be included in the field of aktionsarts. Thus, these suffixes not only add new shades of meaning to their verbal stems, but may also vary the character of action, if it is realized as a process. Certain derivational suffixes were very productive, so the Manchus could easily form new verbs by adding them to nominal words, adverbs, and verbs proper.
gorel-1.p65
234
4/17/02, 8:57 AM
235
morphology
The following is a list of productive suffixes, most of which were used to derive new verbs from nominal stems, although some could also be attached to verbs: 1) -la/-le/-lo: aba “hunt,” “battue”—abala- “to hunt;” gucu “friend”— gucule- “to make friends,” “to be friends with;” akdun “firm,” “strong”—akdula- “to protect,” “to defend;” ejen “ruler,” “lord,” “master”—ejele- “to rule,” “to be master of,” “to establish control over;” erun “torture,” “punishment”—erule- “to torture,” “to punish;” genggiyen “bright,” “clear”—genggiyele- “to make clear,” “to make bright,” “to elucidate;” hahi “urgent,” “hurried”—hahila- “to act quickly or urgently,” “to hurry;” hafu “penetrating,” “going through,” “through”—hafu- “to penetrate,” “to go through,” “to pierce”— hafula- “to penetrate,” “to go through;” lasha “asunder,” “into sec tions,” “in two,” “into pieces”—lashala- “to break off,” “to break in two;” oron “vacant post,” “vacancy”—orolo- “to fill in,” “to fill a vacancy,” “to put a substitute in;” sadun “a relative by marriage”— sadula- “to form an in-law relationship;” sarin “banquet,” “feast”— sarila- “to hold a banquet;” songko “trace,” “track,” “footprint”— songkolo- “to follow in the tracks of,” “to imitate;” suhe “ax”—suhele“to split with an ax;” untuhun “empty,” “emptiness”—untuhule- “to be empty.” It should be mentioned that the final -n of the noun stem is de leted. Some verbal stems which have the same derivational suffix -la/-le/-lo are likely to follow the same word formation pattern, but they are not used without this suffix: weile- “to work;” dahala“to follow,” “to pursue;” burula- “to flee,” “to take flight.” 2) -ra/-re/-ro: amba “big,” “great”—ambara- “to do on a large scale;” colgon “peak,” “high promontory”—colgoro- “to surpass,” “to excel;” ehe “bad,” “evil”—ehere- “to become evil or fierce,” “to be on bad terms with someone;” gisun “speech”—gisure- “to speak;” monggo “Mongolia,” “mongolian”—monggoro- “to speak Mongolian,” “to act in a Mongolian manner.” The same suffix is also used to form new verbs from verbal stems: saci- “to chop,” “hack,” “to chop off”—sacire- “to hack,” “to chop at,” “to chop into pieces.” There are some nouns, the verbs from which are derived by both suffixes -la and -ra: gohon “a hook;” goholo- “to hook,” “to put on a hook;” gohoro- “to bend,” “to form the shape of a hook,” “to curl;” taji “naughty,” “mischievous;” tajila- , tajira- “to act naughtily.”
gorel-1.p65
235
4/17/02, 8:57 AM
236
part four
3) the suffix -na/-ne/-no denotes the internal capacity of an object for self-development or self-reproducing: abdaha “a leaf”—abdahana“to leaf,” “to produce leaves;” bongko “bud of a flower”—bongkono“to form a bud;” eifun “a boil,” “a pimple,” “a swelling”—eifune- “to develop a boil or a swelling;” suihe “an ear of grain”—suihene- “to put forth ears (grain); umiyaha “worm”—umiyahana- “to get worms (of fruit).” The same suffix may be attached to verbs and in this case it is normally used to add the semantics of motion, departure from or leaving a place for other places or objects. 4) the suffix -tu is added to nouns to derive verbs: ulin “goods,” “property,” “possessions”—ulintu- “to bribe;” kimun “enmity,” “animosity”—kimuntu- “to feel animosity against smb.” 5) the suffixes -li, -mi are also added to nouns to derive verbs: bolgo “clean,” “clear”—bolgo- “to clean,” “to make clear,” “to clarify,” “to explain”—bolgomi- “to abstain,” “to fast;” doko “lining of a garment”— dokomi- “to line (a garment);” goro “far”—goromi- “to do from afar,” “to go a long distance;” monggon “neck,” “throat”—monggoli- “to wear on the neck;” tohon “button”—tohomi- “to button,” “to button up;” ture “the leg of a boot”—turemi- “to attach the leg of a boot.” 6) the suffix -àa/-àe/-ào denotes permanent efforts which are made to attain a result, duration, and continued actions: adali “like,” “same”—adaliàa- “to resemble,” “to be like;” banuhån “lazy”— banuhåàa- “to be lazy;” buleku “mirror”—bulekuàe “to look in a mir ror;” etuhun “strong,” “powerful”—etuhuàe- “use force;” hitahån “nail (of the finger or toe)”—hitahåàa- “to press firmly with a fingernail;” mangga “difficult”—manggaàa- “to have difficulties;” narhån “fine,” “thin”—narhåàa- “to be fine,” “to make thin;” nimaha “fish”— nimahaàa- “to fish;” oncohon “arrogant,” “overbearing”—oncohoào- “to be arrogant,” “to be overbearing;” sain “good,” “well”— saiàa- “to praise,” “to commend;” saman “shaman”—samaàa- “to perform a sha manistic rite;” teifun “cane,” “staff”—teifuàe- “to to walk with a cane or a staff.” It is likely that this suffix is related to the expression of the durative or frequentative meanings of the verb (see Section 5.4). 7) the suffix -da/-de/-do denotes gradual involvement in an action
gorel-1.p65
236
4/17/02, 8:57 AM
237
morphology
and the duration of a process: acuhiyan “slander”—acuhiyada- “to slan der;” balama “mas,” “crazy”—balamada- “to rave,” “to act crazily;” ceku “a swing”—cekude- (cekule-) “to swing in a swing;” dufe “disso lute,” “lascivious”—dufede- “to act dissolutely;” eruwen “drill,” “auger”—eruwede- “to drill (a hole),” “to make a hole with an auger;” goho “elegant,” “dainty,” “fop,” “dandy”—gohodo- “to adorn oneself,” “to make up;” hanja “honest,” “clean,” “pure”—hanjada- “to act honestly;” hålhi “muddleheaded,” “confused”—hålhida- “to act in a confused manner,” “to be in a daze;” jali “plot,” “intrigue”—jalida“to plot against,” “to intrigue against;” jili “anger,” “temper”—jilida“to get angry,” “to become mad;” onco “broad,” “wide,” “generous”—oncodo- “to forgive,” “to grant amnesty;” oshon “cruel,” “bru tal,” “tyrannical”—oshodo- “to be cruel,” “to be brutal;” yobo “fun,” “play” “joking”—yobodo- “to have fun,” “to joke.” 8) the suffix -ta/-te/-to is also added to nouns to derive verbs: gosin “love,” “mercy,” “pity”—gosita- “to love,” “to like;” tuksin “throb bing,” alarm,” “anxiety”—tuksite- “to be alarmed,” “to be greatly anxious” (the final -n of noun stems is deleated). The meaning of the derivational suffixes -da/-de/-do and -ta/-te/ -to became the semantical base on which they were transformed into aspectual ones. Having undergone further development, when at tached to verbs, they denote the iterative meaning (see Section 5.4). 9) the suffix -ja/-je/-jo in most cases denotes duration or intensity of the mental or emotional process which a person experiences: golohon “fright”—golohonjo- “to be exceedingly frightened;” gånin “thought”— gåninja- “to think over carefully,” “to reflect upon;” ilin “standing”— ilinja- “to stand unsurely (of a small child),” “to loiter,” “to stand around,” “to make a stop;” murin “stubborn,” “stubbornness”— murinja- “to be stubborn,” “to act recalcitrantly;” niyeniyehun “weak willed,” “lacking initiative”—niyeniyehunje- “to be weak willed,” “to be weak of character,” “to lack initiative;” talihån “undecided,” “vacillating”—talihånja- “to vacillate,” “to be undecided;” urgun “joy,” “felicity,” “happiness”—urgunje- “to rejoice,” “to be glad.” Attached to verbs, this suffix has an aspectual or reflexive mean ing. The aspectual meaning is mostly related to the expression of the durative or frequentative meanings (see Sections 5.4 & 5.5). The following suffixes are mostly used in the derivative processes
gorel-1.p65
237
4/17/02, 8:57 AM
238
part four
forming new verbal lexemes from verbs. It is plausible that in some cases the newly formed verbs are related to the expression of the causative meanings: 10) -niye: ebe- “to wet,” “moisten”— ebeniye- “to wet thoroughly,” “to soak” (Norman gives the following lexical meanings of these verbs: ebe- “to become soaked through,” “to become soggy”—ebeniye- “to soak”); ekiye- “to diminish,” “to be deficient,” “to be too little,” “to be lacking”—ekiyeniye- “to lessen,” “to diminish;” 11) -kiya/-kiye, -giya, -hiya/-hiye: aliya- “to wait,” “to regret”—aliyakiya“to linger,” “to pace back and forth while waiting;” bodo- “to calcu late,” “to figure;” bodo-n “calculation,” “plan;” bodonggiya- “to talk to oneself,” “to reckon upon” (in this case it is likely that the suffux -kiya is used to derive the new verb from the noun “bodon”); foso“to shine,” “to light up”—fosokiya- “to get upset because of impa tience;” hafu- “to understand thoroughly,” “to comprehend”— hafukiya- “to inform in detail,” “to give a thorough rundown on;” jalu- “to be full,” “to be fulfilled”—jalukiya- “to fill out,” “to fill up;” necin “level,” “flat,” “peaceful,” “calm”—necihiye- “to level,” “to calm dawn,” “to pacify” (in this case it is likely that the suffix -hiye is ised to derive the new verb from the noun); sesule- (sesula-) “to be startled” —sesukiya- “to shiver from the cold;” taci- “to learn,” “to study,” “to get used to”—tucihiya- “to instruct,” “to train;” tuwa- “to look,” “to look at,” “to observe”—tuwakiya- “to watch,” “to guard,” “to watch over,” “to observe;” yaru- “to lead,” “to guide”—yarkiya- (instead of yaru-kiya-) “entice,” “to lure;” àuàu àaàa “whispering,” “speaking softly;” àuàunje- “to whisper;” àuàunggiya- “to whisper,” “to hum.” A specific source of verbal derivation in Manchu is onomatopoeic words followed by the form of the imperfect converb from the verb se- “to say,” “to call,” “to mean,” i.e. seme “speaking.” Partly dese mantisized, this verbal form takes upon itself a governing function and may be followed by all verbal suffixes. The lexical meaning is rendered by the onomatopoeic words. Here are some examples: bubu baba seme “mumbling”—bubu baba se- “to speak mumbling;” ek seme “a typical sound of someone who has the urge to vomit”—ek se- “it makes someone sick,” “to be annoyed with,” “to be tired of;” hiyor hiyor se- “to be robust;” kaka kiki seme “the sound of happy laughter”— kaka kiki se- “to laugh for one’s joy;” kek se- “to be pleased,” “to be
gorel-1.p65
238
4/17/02, 8:57 AM
239
morphology
gratified,” “to be refreshed;” kes seme “the sound as if cut off with a knife, sharp, sheer”—kes se- “to cut of with a knife or a sickle.” Used frequently with the converb seme, certain onomatopoeic words were amalgamated with the verb stem se-. The latter was changed into the suffix -si or -sa, following the stem morpheme. That became the base from which new verbs could be derived. The following verbs have developed as a result of this process: kabkasa- “to answer impudently,” “to talk back to” (kab seme “snap ping, beating (of a pack of dogs);” labsi- “to gulp down,” “devour” (lab seme “with the mouth packed full”); longsi- “to chatter on and on,” “to talk foolishness” (long seme “to keep on talking foolishly,” “to talk nonsense”); tuksi- “to pound, to throb (of the heart)” (tuk tuk seme “pounding of the heart”). Being part of the verbal stem, derivational suffixes always precede the suffixes of aspect, voice, and mood (tense and modality). 5.3. Verbs of Motion In Manchu there are two derivational suffixes denoting two types of motion, opposed to each other. One of the types is connected with the idea of departure from or leaving a place for other places or towards other objects of an action, the other one expresses the idea of coming or arrival to other places as well as the achievements of the agent’s object. Followed by these suffixes, in addition to their basic semantics verbs take on the meaning of motion. The suffix -na/-ne/-no which is most likely historically connected with the verb gene- “to go” adds to verbs the meaning “to go (to do smth.): ala- “to tell,” “to report”—alana- “to go to report;” ebiàe- “to bathe,” “to swim”—ebiàene- “to go to bathe or swim;” fata- “to pinch,” “to pick (fruit)”—fatana- “to go to pinch or pick;” feku- “to jump,” “to leap”—fekune- “to jump across (away from the speaker),” “to jump to the other side;” fude- “to see off”—fudene- “to go to see off;” guri“to move,” “to transfer”—gurine- “to go to move,” “to move to another place;” hengkile- “to kowtow,” “to prostrate oneself”— hengkilene- “to go to kowtow,” “to go to court;” okdo- “to go to meet,” “to greet”—okdono- “to go out to meet,” “to go to greet;” omi- “to drink”—omina- “to go to drink;” simne- “to examine,” “to take an examination,” “to test”—simnene- “to go to examine,” “to go to take an examination;” te- “to sit,” “to reside”—tene- “to go to sit,” “to go to reside.”
gorel-1.p65
239
4/17/02, 8:57 AM
240
part four
Followed by this suffix, verbs denote depature (going out) of the agent from a place to other places or towards the object of the action. This synthetic form can be replaced by an analytical one, the first component of which is the form of the imperfect converb of an autosemantic verb and the second is represented by one of the forms of the verb gene- “to go:” alana-, alame gene- “to go to tell,” “to go to report;” tene-, teme gene- “to go to sit,” “to go to reside.” The suffix -nji which is probably derived from the verb ji- “to come” adds to verbs the meaning “to come (to do smth.)”: ala- “to tell,” “to report”—alanji- “to come to report;” fata- “to pinch,” “to pick (fruit)”—fatanji- “to come to pinch or pick;” feku- “to jump,” “to leap”—fekunje- “to jump across (toward the speaker);” fude- “to see off”—fudenji- “to come to see off;” guri- “to move,” “to transfer”— gurinji- “to come to move,” “to move here;” hengkile- “to kowtow”— hengkilenji- “to come to kowtow,” “to come to court;” okdo- “to go to meet,” “to greet”—okdonji- “to come to meet;” omi- “to drink”—ominji“to come to drink;” simne- “to examine,” “to take an examination”— simnenji- “to come to examine,” “to come to take an examination;” te- “to sit,” “to reside”—tenji- “to come to sit,” “to come to reside.” Followed by this suffix, verbs denote coming, arrival of the agent to the place of the action or achievement of the agent’s object. This synthetic form can be replaced by an analytical one, the first com ponent of which is the form of the imperfect converb of an autono mous verb and the second is represented by one of the foms of the verb ji- “to come:” alanji-, alame ji- “to come to tell;” tenji-, teme ji“to come to sit,” “to come to reside.” The verb ji- “to come” has an irregular form of the imperative, i.e. jio. When followed by the suffix -nji, verbs in the imperative also have irregular forms: alanju “come and tell;” dosinju “come in;” okdonju “come to meet” (Zakharov, 1879:165). 5.4. Aspect The grammatical category of acpect needs to be further investigated. When scholars discuss the problem of aspect in the Manchu language they take into consideration first of all the imperfect and perfect participles formed by the suffixes -ra/-re/-ro and -ha/-he/-ho respec tively. According to the Russsian tradition in Manchu studies, the first form denotes the present-future and the second is connected with the expression of the past. In fact, these two participles denote
gorel-1.p65
240
4/17/02, 8:57 AM
241
morphology
aspectual meanings rather than temporal. The participle in -ra/ -re/-ro denotes the imperfect aspect in the grammatical space of the present-future tense, and the participle in -ha/-he/-ho conveys the perfect aspect in the grammatical space of the past. As shown by V.A. Avrorin, when these two forms were fixed in classical Manchu, the aspectual meanings of both of them were in the process of transformation into temporal meanings (Avrorin, 1949). Opposition between these two participles is of great importance not the least because they play an important role in the structure of the Manchu utterance. It is not an overstatement to say that the two participles are the basis from which most finite verbal forms have developed. Apart from the grammatical opposition between the imperfect and perfect aspects which is limited to participles, in classical Manchu there are a few synthetic forms, albeit not as many as in other Tungusic languages, which characterize an action from the point of view of quality and quantity. Expressed by suffixes, these forms display not only different shades of lexical meanings but also aspectual distinctions in the verb. On these grounds they can be seen as in stances of aktionsarts, in the sense this term is used by scholars work ing in Slavonic languages, rather than aspect markers (Agrell, 1908; Noreen, 1904-1912; Seidel, 1936; Isachenko, 1960; Maslov, 1962; Bondarko & Bulanin, 1967; Bondarko, 1971). In some cases it is doubtful whether suffixes specified as aspectual are actually effect verbs respecting certain aspectual distinctions or whether they are part of verbal stems as lexical entries, i.e. can be seen as derivational ones. Another difficulty lies in the fact that certain suffixes may denote aspectual meaning with some verbs, but with others display the meaning of voice (in the broad sense). It is questionable whether these suffixes are the same, and if so, what generic semantics might be the basis of these two lexical-grammatical categories. Denoting aspectual distinctions in the verb, most suffixes may combine a series of meanings. Each of these meanings is realized depending on the lexical semantics of a concrete verb. The combi nation of meanings is limited by the semantics of iteration, frequency, duration, continuity, or intensity of an action. Most frequentative verbs are formed by these suffixes. The adverbs “repeatedly,” “per manently,” “constantly,” and “intensively” most adequately describe the semantics of these newly derived verbs.
gorel-1.p65
241
4/17/02, 8:57 AM
242
part four
The suffix -àa/-àe/-ào has the durative meaning with most verbs, but with some it displays the frequentative meaning: jafa- “ to catch,” “to hold,” “to take hold of,” “to grasp,” “to grip,” “to seize”—jafaàa“to hold continually,” “to hold for a while,” “to keep groping for;” kurbu- “to turn around,” “to turn over,” “to toss and turn (while sleeping)”—kurbuàa- “to turn round and round,” “to roll over and over;” “to be upset,” “to be enxiuos;” nara- “to long for,” “to feel attached to,” “to linger over”—naraàa- “to long for continually,” “to linger over a long time;” olho- “to fear”—olhoào- “to be careful,” “to be cautious.” When the suffix -àa/-àe/-ào is added to the verbs ending with the suffix -ra/-re/-ro the vowel of the latter is dropped: fatara- “to pinch repeatedly,” “to use things sparingly”—fataràa- “to pinch repeatedly,” “to be in the hebit of using things sparingly;” hafira- “to pinch,” “to press or hold between two objects,” “to put pressure on,” “to compel”—hafiràa- “to live economically,” “to use economically;” sefere“to take a handful,” “to grab with the hand”—sefereàe-, seferàe- “to keep on taking handsful,” “to keep on grabbing;” àoforo- “to pinch,” “to take a pinch of,” “to scratch”—àoforào- “to scratch all over,” “to take random pinches of.” According to Zakharov, when the suffix -àa/-àe/-ào is attached to the verbs ending in the suffix -la/-le/-lo the latter is deleted: habtala“to squint,” “to wink,” “to blink”—habtaàa- “to wink or blink repeat edly;” hengkile- “to kowtow”—hengkiàe- “to kowtow repeatedly;” àasihala- “to slap,” “to clap”—àasihaàa- “to slap repeatedly” (Zakharov, 1879:166). It is obvious that this suffix, displaying in the examples cited above, the aspectual meaning is closely related to the derivational suffix -àa/-àe/-ào which serves to form new verbs from nouns. These de rivative verbs also denote durative and continuous actions (see Sec tion 5.2). Preceded by additional elements -n, -l,-r, or -gan, -hon, -hun, the suffix -ja/-je/-jo when used with some verbs, has the aspectual mean ing which is mostly durative. In some cases this suffix denotes the intensity of an action. Used with the following vebs, this suffix adds to them the meaning “to do smth. constantly or intensively:” ere- “to hope”—erehunje- “to hope constantly;” golo- “to be startled,” “to be scared”—golohonjo- “to be exceedingly frightened;” gåni- “to think,” “to reflect,” “to consider,” “to intend”—gånunja- “to think over carefully,” “to reflect upon;” guri- “to move,” “to transfer”— gurinje
gorel-1.p65
242
4/17/02, 8:57 AM
243
morphology
“to move from one place to another constantly,” “to lead a nomadic life;” fulara- “to become red,” “to blush”—fularja- “to have a red ap pearance;” haya-”to coil,” “to twist”—hayalja- “to wind constantly,” “to twist constantly.” With other verbs the suffix -ja/-je/-jo has mostly reflexive mean ing and should be analyzed in the chapter devoted to the grammatical category of voice (see Section 5.5). The suffix -ja/-je/-jo is used to form new verbs from other parts of speech. These verbs also denote certain degree of making efforts, intensity or continuity of an action or process (see Section 5.2). The suffix -ca/-ce/-co has the frequentative meaning. With some verbs it may denote duration or intensity of an action. The follow ing verbs display the additional aspectual meaning which is formed by this suffix: feku- “to jump,” “to leap”—fekuce- “to leap up,” “hop over;” hira- “to look askance at,” “to spy on”—hiraca- “to keep looking askance,” “to spy on intently;” jolho- “to gush up,” “to well up”jolhoco- ( jolgoco-) “to stampede,” “to rush off in a fury,” “to press forward in a rage.” When this suffix is added to verbs ending in the suffixes -ra/-re/ -ro or -la/-le/-lo, the vowel of the latter is deleted: bedere- “to with draw (at court or at a ceremony)”—bederce- “to withdraw many times;” sosoro- “to back up,” “to withdraw”—sosorco- “to back up,” “to shrink back.” Norman gives the following translations of the pair of verbs: bedere- “to return”—bederce- “to retreat.” These meanings cannot be considered as aspectual (NL:26). The suffix -ca/-ce/-co is used to denote the associative meaning which should be analyzed in the chapter devoted to the grammati cal category of voice (see Section 5.5). To denote the iterative meaning the suffix -ta/-te is used. With some verbs it displays the meaning of continuous action. The suffix -ta/-te can be found, for example, in the following verbal stems: ana“to push”—anata- “to push repeatedly;” debsi- “to fan,” “to flap,” “to flutter”—debsite- “to fan continually,” “to flutter continually;” fehu“to step on,” “to tread on,” “to trample”—fehute- “to trample repeat edly;” jafa- “to take in the hand,” “to hold,” “to grasp,” “to grip”— jafata- “to grasp or to grip repeatedly” (Norman gives another trans lation of the verb jafata-, which is “to keep in rein,” “to restrain;” NL:153); niyece- “to mend,” “to fill (a post),” “to fill in”—niyecete- “to mend continually,” “to fill in regularly;” tuksi- “to pound,” “to throb (of the heart),” “to be alarmed,” “to be anxious”—tuksite- “to throb
gorel-1.p65
243
4/17/02, 8:57 AM
244
part four
(of the heart) continually,” “to be greatly anxious;” uàa- “to pull” - uàata- “to pull with force.” The verb stem tuwaàata- “to look at smth. or smb. for a long time and repeatedly,” “to look after” contains two aspectual suffixes one of which is used to denote the durative meaning (-àa), and another is used to denote the iterative meaning (-ta). This verb was derived from the verb tuwa- “to look,” “to look at.” 5.5. Voice The grammatical category of voice is one of the most discussed topics in general linguistics. However, there is currently no conventional definition of voice. In Tungus-Manchu studies voice also remains one of the most problematic grammatical categories. The seman tics of voice is still not quite clear, and scholars continue to ascribe semantically different forms to this category. Many scholars maintain that voice doesn’t possess any semantics, and that the main function of this category is to transform the syn tactic structure of a sentence, i.e. the subject and the object switch roles, and this fact is reflected in the verbal form. The universal theory of voice was developed in the early seven ties. In this theory, the concept of semantic and syntactic category of diathese is used, i.e. the correspondence between the semantic roles of a verbal lexeme (agent, patient, addressee, etc.) and its syn tactic roles (subject and object). The latter coincide with the parts of the sentence. Such an understanding of the category of diathese can be found in Tesniere’s Elements de Syntaxe Structurale (Tesniere, 1976). Voice is defined as a morphologically marked diathese. In other words, voice is a mechanism, which permits a change of the diathese of a verb lexeme. The opposition between the active and passive forms underlies the category of voice in many languages. The verbal form is active when the subject corresponds with the agent while in the passive voice the subject corresponds with the patient (Mel’chuk & Kholodovich, 1970; Kholodovich, 1979; Khrakovsky, 1979; LED, 1990). Other scholars consider the category of voice as closely related with the communicative structure of a sentence. Such an understand ing of the category of voice has recently been put forward by V.A. Plungyan in his “General Morphology” (Plungyan, 2000:191-224). Plungyan believes that two classes of forms should be distinguished in the semantic space of voice. The first class comprises forms of voice
gorel-1.p65
244
4/17/02, 8:57 AM
245
morphology
proper, while the second one includes forms of actant derivation. The category of voice has its own semantics, albeit this semantics is unique. The forms of voice express the attitude of the speaker to the information that is announced. Such semantic elements are referred to as communicative or pragmatic information. The mor phological verbal forms of voice (as a rule, together with syntactic devices) are used to express communicative and/or pragmatic op positions. Voice can be defined as a verbal category, whose forms are used to indicate certain changes of the communicative rank of participants who are engaged in a situation of the outside world. The active voice indicates the initial (basic) rank structure of a sentence. The oblique voices are used to transfer the status of a participant with the high est communicative rank from one argument to the others. In fact, the function of voice is not so much the redistribution of syntactic roles as the redistribution of communicative ranks. If a verb has two or more actants (the parts of the sentence that indicate participants of a situation), one of them is opposed to the others as the subject vs. objects. Being the principal actant in many languages, the subject is characterized by the highest communica tive rank (topicality). In order to decrease the communicative rank of a subject, the forms of passive voice are used. In canonical pas sive constructions, the communicative rank of a direct object is normally increased, and the object occupies the position of a sub ject. The communicative rank of the subject is decreased, and the subject occupies the position of the object. However, there are passive constructions in which the status of the patient is not increased. Passive constructions with “zero-agent” also exist. The opposition of “active vs. passive” is considered central to the category of voice. What is very important to note is that the situation, which is described with the help of the passive construction, remains the same when the voice transformations occur. In the theory of voice accepted in Altaic studies, the category of voice is represented by the opposition between active and passive constructions. In the active voice the agent is placed in the position of a subject and rendered by the nominative. In Manchu as in the other Altaic languages the active voice has no special morphologi cal marker and is semantically and morphologically opposed to the passive. The active comprises the transitive and intransitive verbs: ala- “to tell,” te- “to sit,” “to live.” To manifest the passive voice the
gorel-1.p65
245
4/17/02, 8:57 AM
246
part four
suffix -bu is used. This suffix belongs to a verbal stem and corre spondingly to all verbal forms, which are derived from this stem. All verbs, which do not have this suffix, can be seen as active. In the passive voice the subject is placed in the position of object and ren dered by one of the oblique cases, dative in the case of Manchu: bi in-de gele-bu-he I he(i/in-)-DAT fear-PASS-PART “I was frightened by him” (PASH:34); in-de gida-bu-ha he(i/in-)-DAT press-PASS-PART “(Someone) was pressed by him;” baita de uàa-bu-ha matter DAT burden.with-PASS-PART “(Someone) was burdened with the matter” (ZAKH:160); weri de basu-bu-mbi, gånin de somebody.else DAT deride-PASS-IMPF thought DAT gida-bu-mbi hurt-PASS-IMPF
“(One) is derided by strangers and hurt by thoughts (of others)”
(PASH2:57).
The object takes the position of subject in the passive construction
and is marked by the nominative.
To form the passive voice the suffix -mbu is also used with some verbs. According to Zakharov, this suffix has a different shade of meaning in contrast with the suffix -bu (Zakharov, 1879:160). The passive voice can be expressed analytically, by the verb isibu“to bring to,” “to deliver,” “to send,” “to take to,” “to reduce to,” “to drive (into, to)” which is originally the causative form of the verb isi- “to reach,” “to arrive;” “to approach,” “to come up to;” “to suffice;” “to be about to (with the imperfect participle):” in-i arga de buce-bu-he he(i/in-)-GEN plot DAT die-PASS-PART “(Someone) was done away by his plots;” in-i arga-i buce-re de isi-bu-ha he(i/in-)-GEN plot-GEN die-PART DAT reach/approach/ be.about-CAUS-PART “(Someone) was driven to death by his plots” (ZAKH:160).
gorel-1.p65
246
4/17/02, 8:57 AM
247
morphology
Some passive verbs, derived from active ones by the suffix -bu, take the reflexive meaning. Constructions with these verbs do not con tain the NP (nominal phrase) which occur in the dative to express the agent in passive constructions: gurun dasa-bu-mbi state/ruling.house rule/correct/repair-PASS-IMPF “The state improves;” irgen beye dasa-me mute-rakå dasa-bu-re people self rule-CONV be.able-PART(NEG) rule-PASS-PART de akda-mbi DAT trust-IMPF “People are unable to govern themselves, therefore (they) entrust (themselves) to rulers” (ZAKH:160). As Plungyan mentions, the notion of the so-called actant derivation is not widely accepted. This notion is associated with the term “verb deriving morphology,” introduced by B. Comrie (Comrie, 1985:30148). The principal distinction between voice and actant derivation is that voice transformations change the pragmatic interpretation of a situation, albeit never affect its semantic interpretation. On the contrary, it is changes in the number and/or referential properties of participants of a situation, i.e. semantic transformations of the basic structure, that are distinctive features of actant derivation. The re distribution of communicative ranks also occurs (as a rule), albeit this change is not the only one. In a sense, it is possible to think that the category of voice is a particular case of the category of actant deri vation (Plungyan, 2000:208). Actant derivation is called increasing (the so-called “valency increasing derivation” according to Comrie) when a new obligatory participant in comparison with the basic situation appears, and this change is reflected in the morphological structure of a verb. The appearance of a new participant with the role of agent occurs more frequently, and in this case it is accepted to say about derivative causative verbs. It is very important to note that this new partici pant always occupies a privileged syntactic position. The new agent becomes a subject (consequently the previous subject decreases in topicality). In Manchu, the polyfunctional suffix -bu is also used to express the causative. In comparison with the active construction, in the
gorel-1.p65
247
4/17/02, 8:57 AM
248
part four
causative one the subject is transformed into an agentive object which is the real performer of the action (agent). This action is performed according to the will (or desire, permission, assistance, order, etc.) of the second person who is the causator. The latter causes the agent to perform the action rendered by a verbal stem. In causative con structions the subject is the causator of an action, but not the real agent. The agent is expressed by a word in the accusative: tere-be gene-bu this(he)-ACC go-CAUS(IMP) “Order (tell) him to go;” i mim-be gele-bu-he he I(bi/min-)-ACC fear-CAUS-PART “He caused me to be frightened;” tere-be baita icihiya-bu-me gene this/he-ACC matter arrange-CAUS-CONV go(IMP) “Go and order him to arrange the matter” (ZAKH:160); amba-sa saisa be an be tuwakiya-bu-mbi high.official-PL gentleman ACC usual ACC keep-CAUS-IMPF buya niyalma be waka be ulhi-bu-mbi ordinary person ACC mistake ACC understand-CAUS-IMPF ambasa saisa “a wise man,” “true gentleman” (NL:15); “(They) force true gentlemen to follow what is customary, and force ordinary persons to realize mistakes” (PASH2:54); niyalma be sain baita yabu-bu-mbi person ACC good affair/matter do/perform-CAUS-IMPF “(They) make persons to do good things” (PASH2:55). Here are some verbs in the form of the passive voice and the caus ative: wa- “to kill”—wabu- 1) “to order smb. to kill,” 2) “to be killed;” weile- “to work,” “to make”—weilebu- 1) “to order smb. to work,” 2) “to be made;” karula- “to repay,” “to recompense”—karulabu- 1) “to cause (repayment or smb. to repay),” “to cause (recompense or smb. to recompense),” 2) “to be repaid,” “to be recompensed.” The fact that the same suffix -bu is used to express the passive and causative meanings, whose syntactic functions are opposed to each other, is rather puzzling. Although there are some more languages where the polysemy of the passive and causative markers is noted (for example, Korean, a number of Turkic, etc.), the passive more
gorel-1.p65
248
4/17/02, 8:57 AM
249
morphology
frequently developed from the markers of decreasing derivation, as a result of their grammaticalization (Plungyan, 2000:220). Normally verbs followed by the suffix -bu take part in syntactic constructions that have passive and causative semantics. In some cases, however, this suffix is used to form the causative pair of verbs and performs a derivational function: be- “to be,” “to exist;” “to stay,” “to remain”—bibu- “to detain,” “to keep back,” “to retain,” “to leave behind;” sa- “to know,” “to understand”—sabu- “to see,” “to per ceive.” Newly derived verbal stems, in their turn, also may take the suffix -bu: sabubu- PASS/CAUS of sabu-. With some verbs the suffix -mbu is used to form the causative: dosi- (the perfect participle has the form in -ka) “to enter,” “to advance”—dosimbu- “to order smb. to enter, to advance;” suru- (the perfect participle has the form in -ke) “to quiet down,” “to calm down”—surumbu- “to order smb. to quiet down.” Here is an example with the causative suffix in the form -mbu: niyalma be gånin usa-mbu-mbi, jaka be person ACC intention be.disappointed.in-CAUS-IMPF thing ACC teisu baha-bu-mbi assigned.place get-CAUS-IMPF “(They) force a person to be disappointed, and force a thing to get an assigned place” (PASH2:54). The suffix -mbu has a different shade of meaning in contrast with the suffix -bu. Here are two verbs which may be followed by both of these suffixes: wasi- (the perfect participle has the form in -ka) 1) ”to descend,” “to go down,” “to sink;” 2) ”to fall (of rulers);” 3) “to decline (of value)”—wasibu- 1) CAUS of wasi- ; 2) “to demote,” “to degrade”—wasimbu- 1) CAUS of wasi- ; 2) ”to demote,” “to degrade;” 3) ”to issue (an order),” “to send down (an edict);” wesi- (the perfect participle has the form in -ke) 1) ”to ascend,” “to go up,” “to raise;” “to advance (in rank)”—wesibu- 1) CAUS of wesi- ; 2) “to lift,” “to raise;” 3) “to promote,” “to advance”—wesimbu- 1) CAUS of wesi- ; 2) “to raise,” “to lift;” 3) “to advance,” “to promote;” 4) ”to submit, to present (to the emperor),” “to report to the throne” (Zakharov, 1879:160; NL:305, 308). The materials of Norman and Yamamoto display the fact that in Sibe the suffix -v� is used to denote the causative and passive. This fact may be considered as more evidence on the influence of the
gorel-1.p65
249
4/17/02, 8:57 AM
250
part four
Tungusic languages on Sibe because the suffix -v� corresponds to the passive suffixes in the Negidal, the Even, and the Evenki lan guages: neg. -v, -mu; even. -v, u/-u, -m; evenk. -v, -b ~ -p ~ -mu (Konstantinova, 1964:154-5; Novikova, 1960:55-6; Tzintzius, 1982:23). The Manchu suffix -nggi, although seldom used, has a meaning close to that of the causative. This suffix has the meaning “to send smb. to do smth.:” alanggi- “to send to tell, to report” (ala- “to tell,” “to report”); boolanggi- “to send to report” (boola- “to report”); tuwanggi“to send to look,” “to send to examine” (tuwa- “to look,” “to look at,” “to examine”). To denote the same meaning the verb unggi- “to send” is usually used instead. For example, the verb alanggi- “to send to tell” can be replaced by the analytical construction with the same meaning: alame unggi- “to send to tell” (ala-me is the form of the imperfect converb of the verb ala- “to tell,” “to report”). There is one more type of increasing actant derivation which is very important for the Manchu language. The associative suffixes -nu, -ca/-ce/-co are used to mark not the appearance of a partici pant with a new role, but the appearance of a new participant or participants with the same role. These suffixes indicate that an ac tion is performed by several participants together: afanu- “to attack together,” “to fight together” (afa- “to attack,” “to fight”); amgaca“to sleep together” (amga- “to sleep”); deduce- “to lie down together,” “to sleep together” (dedu- “to lie,” “to lie down”); dosinu- “to enter together” (dosi- “to enter”); ilica- “to stand together” (ili- “to stand,” “to spot”); kicenu- “to strive together” (kice- “to strive”); omica- “to drink together” (omi- “to drink”); songgoco- “to weep together” (songgo- “to weep”); tece- “to sit together,” “to sit down together” (te- “to sit,” “to live”). According to its semantics, the associative takes interim place between the area of actant derivation and the area of verbal plural ity. The associative meaning can be expressed by words which de note plurality: gemu “all,” geren “a crowd,” “many,” “many kinds of.” Therefore instead of the synthetic form with the associative suffix, the combination of one of these words and a verb can be used: isanu-ha and geren isa-ha “all gathered together;” dosi-nu-ka and geren dosi ka “all entered.” All verbs that have the associative suffix may produce causative and passive forms. The only difference is that the causative/passive suffix -bu is normally followed by the suffix -nu, but it is preceded
gorel-1.p65
250
4/17/02, 8:57 AM
251
morphology
by the suffix -ca/-ce/-co: dosibunu- “to order all of them to enter to gether;” tecebu- “to order all of them to sit together” or “to be seated all together.” The imperative of the causative and passive forms of the associa tive verbs may be expressed by a combination, the first component of which is represented by an autonomous verb and the second being the imperative form of the functional verb se- “to say:” dosibunu or dosinu se “order all of them to enter together”9. Decreasing actant derivation (the so-called “valency-decreasing derivation” according to Comrie), on the contrary, is related with the elimination of one of the actants from a situation. As a result, a new situation of the outside world is created. The most widespread type of decreasing actant derivation is the decausative one, which is, according to its semantics, reverse to the causative. In Manchu, the suffix -ja/-je/-jo expresses the decausative mean ing with some verbs, although sometimes it can mark the reflexive and passive as well. When this suffix is attached to them, the fol lowing verbs display these meanings: debkeje- “to come loose,” “to come unraveled” (debke- “to bring up again,” “to take up again;” debkele- “to untwist,” “to unravel”); dokdorja- “to act unsettled,” “to be erratic,” “to jump about;” efuje- “to be ruined,” “to be defeated,” “to be dismissed from a position” (efule- “to destroy,” “to ruin,” “to break,” “to dismiss”); fondojo- “to be broken or torn through” ( fondo-, fondolo- “to penetrate,” “to go through”); fudeje- “to develop a flaw,” “to crack,” “to rip;” halanja- “to exchange in turn,” “to take turns” (hala- “to exchange,” “to change (clothing),” “to take the place of”); kambulja- “to be soft and damp,” “to be swampy;” lumburja- “to be soft, not firm (of wet earth);” sunggelje- “to shake,” “to tremble” (sungge“to waste away,” “to grow skinny and pale”); uyalja- “to move winding like a snake,” “to slither.” The reciprocal suffix -ndu is used to express a particular case of relationships between participants of a situation. It expresses an action performed through a mutual interaction of one subject with another. The number of participants does not reduce to either of the two. However, each participant takes upon him/herself the role of the other participant. As a result, each participant performs his/her own and the other one’s roles simultaneously. 9 Zakharov noted the fact of the irregular use of the suffix -nu instead of the suffix -na/ne/-no which denotes the meaning “to go to do smth.” in the impera tive: dosinu instead of dosina “come in,” tucinu instead of tucine “go out,” “get out,” tucibunu instead of tucibune “order smb. to go out” (Zakharov, 1879:170).
gorel-1.p65
251
4/17/02, 8:57 AM
252
part four
The adverb ishunde “mutually,” “to one another” is the key word used to describe the semantics of the reciprocity. Here are some verbs in the reciprocal form: afandu- “to fight against each other” (afa- “to fight,” “to attack”); aisilandu- “to help one another” (aisila- “to help”); huwekiyendu- “to rouse one another” (huwekiye- “to rouse oneself,” “to be enthusiastic”); jurcendu- “to oppose one another” (jurce- “to dis obey,” “to go against,” “to go against one’s word,” “to turn the back on,” “to oppose in battle”); sirandu- “to follow after one another” (sira“to continue,” “to follow,” “to connect,” “to tie together,” “to in herit”); somindu- “to hide one another” (somi- “to hide”). Sometimes the boundary between the associative and recipro cal meanings is obliterated, and one is used instead of the other. This semantic vagueness also emerges in Norman’s Lexicon where the associative and reciprocal forms are given under a single lexical entry and translated by the adverb “together.” 5.6. Participles The term “participle,” the name used for the most universal and probably most ancient verbal form, is treated by contemporary schol ars conventionally to a certain degree. The conventional character of this term has already been discussed in scientific literature. This term is used to refer to the central and fundamental form of the Altaic (and Manchu) verb, or more precisely, to the hyperform which comprises a few grammatical paradigms on the the basis of a single generic form. Three functions of the Manchu participles are known with which three different grammatical models of participles can be put in correlation. The first is the function of a predicate in a simple sentence or in a principal clause of a complex sentence. In this role participles reveal their verbal characteristics, such as aspect, voice, mood (tense and modality), and person. The ability to govern noun cases is the dis tinctive feature of the Altaic (and Manchu) participle as a verbal form. If a language has the morphological category of person, for in stance, like all Tungusic including Evenki, participles are directly followed by personal suffixes of the predicative type with the help of which they are normally conjugated. Based on the principle of indeclinability and conjugation in this syntactic function, participles have a good case for being specified as verbs proper.
gorel-1.p65
252
4/17/02, 8:57 AM
253
morphology
If a language does not have the morphological category of per son, like Manchu and Mongolian, the predicate correlates to the agent analytically, through the subject. Serving as a principal predicate, a participle takes final position in a sentence in contrast to a participle in the attributive function. Serving as an attribute, a participle always precedes a noun which it modifies. The second function of the participles is that they may serve as attributes or predicates in attributive constructions. It is in this sec ond function that the Altaic (and Manchu) participles may be com pared with their Russian counterparts, and really prove their name. In this function Manchu participles precede nouns which they modify. The morphological form of the participles in the attributive func tion can be defined as indeclinable. The type of syntactic bond between a noun and a participle can be specified as juxtaposition. The following examples show that in the attributive role Manchu participles can neither be declined nor conjugated: na i tuci-ke jaka earth GEN sprout-PART thing
“thing that sprouted from earth;”
bi sin-de yandu-ha baita be si te-de
I you(si/sin- SG)-DAT request-PART matter ACC you this/he-DAT
hendu-hebi-o?
speak-PAST-INT
“Did you speak to him about the matter which I requested of you?”
(PASH:59).
The third function of the Altaic (and Manchu) participles is that they may serve as predicates in subordinate clauses. From the point of view of the Altaic complex sentence theory, this function of par ticiples is of special interest. Particular attention should be paid to the fact that in the role of subordinate predicates Altaic participles may take markers of oblique cases or may be used in the nomina tive case with zero expression. Instead of the term “nominative,” some specialists suggest using the so-called “direct” case which co incides formally with the verbal stem. Participles in most Altaic languages may be followed by personal possessive suffixes. Used with subordinate predicates, possessive suffixes become markers of the predicative-possessive type, and in this way the participles themselves can be conjugated. The ability to be conjugated permits, firstly, the
gorel-1.p65
253
4/17/02, 8:57 AM
254
part four
preservation of the predicative function of the participles within subordinate clauses and, secondly, with the help of cases, the plac ing of participles into noun positions within subordinate clauses. In other words, participles are allowed to function as noun analogues, viz. predicative subjects and objects, both direct and indirect. Manchu participles, having neither the morphological category of person (like the verb proper) nor the morphological category of possession, may be followed only by case markers. Morphologically the Altaic par ticiples in the role of subordinate predicates can be characterized as declinable forms which in most languages may be conjugated. It is precisely these forms that underlie the mechanism of predicative declension of participles which may rightfully be called the basic mechanism of Altaic hypotaxis. Able to take any nominalizing suffixes (case markers and nomi nalizers) participles are often used as substantives. In this function they act as a subject or an object governed by verbal forms. 5.6.1. Imperfect and Perfect Participles In Manchu there are imperfect and perfect participles. These are characterized by a certain temporal meaning—present or past—when they serve as a predicate in a simple sentence or in a principal clause of a complex sentence. In other functions they reveal an aspectual meaning rather than a temporal one (Avrorin, 1949). Imperfect and perfect participles may be followed by case mark ers when they function as nominal parts of a sentence, viz., as predi cates or leading words of predicative constructions which, in their turn, play the part of a subject or an object. Such predicative sub jects and objects are governed by the principal predicate of the sentence. As mentioned above, the nominative case has no special marker. Therefore zero expression of the case should indicate that a word is used in the nominative. This is correct for all nouns which cannot be considered apart from the category of declension. However, in regard to participles, the absence of a case marker cannot be taken for the nominative form. The participle form which acts in two functions in a sentence—as a predicate and an attributive to nouns— should be analyzed outside of the case system. The same out-of-case form is used when the participle functions as a predicate of clauses if they are subordinated to a principal predicate by means of syn tactic words as in:
gorel-1.p65
254
4/17/02, 8:57 AM
255
morphology
tere niru-re fakài ji-he mangi han hendu-me that paint-PART craftsman come- PART after khan say-CONV “After that painter had come the khan said” (SK). Depending on its function in the sentence, the participle may be translated in different ways. When acting as a predicate of a simple sentence the participle is translated as a verb. In the attributive function it is translated as a participle. In the function of a predi cate of a clause, the participle should be translated by a combina tion of a verb and a conjunction. When used as an object the par ticiple should be translated by a verbal noun or a verb which is a predicate of an object clause. The imperfect participle has the suffix -ra/-re/-ro: gene-re (gene- “to go”); ana-ra (ana-”to push”); taci-re (taci- “to learn,” “to study”). There are two irregular participial forms, bisi-re and ojo-ro. The first parti ciple, bisire, is formed from the existential verb bi- “to be,” “to ex ist” which in this case is used in the irregular form bisi-. All the other forms of this verb are formed from the stem bi-. The stem bisi- may be analyzed etymologically bearing in mind the Evenki participle bisi in which -si is an ancient participial suffix. The same suffix can be found in the Evenki negative auxiliary participle esi. Thus, the Manchu participle bisire has two participial suffixes, the first of them is the ancient suffix -si, analogous to the Evenki suffix -si, and the second one is the contemporary suffix -re. The participle ojoro is etymologically equal to the Evenki verb ojara the structure of which is transparent. The verb ojara is the form of the present tense of the third person, plural, of the verb o- “to become,” where -ja is an aspectual suffix. Since this suffix does not occur among living deri vational elements in Manchu, the participle ojoro has a form differ ent from that of other imperfect participles. There are a few more imperfect participles that have irregular forms: jetere (je- “to eat”), jidere (ji- “to come”). A group of imperfect participles have the combina tions of elements n+de inserted between the stem and the suffix -ra/-re/-ro: bandara (ba- “to be tired,” “to be lazy,” “to gnaw a hole”); jendere (je- “to bear,” “to put up with,” “to tolerate”); jondoro (jo- “to bring to mind,” “to recall,” “to mention”). Normally such verbs have irregular forms of perfect participles in -ngka/-ngke/-ngko. According to Iv. Zakharov, some verbs have two forms for the imperfect par ticiples, in -ra/-re/-ro and -ndara/-ndere/-ndoro. The latter form de notes progression of the action: akjara “thundering,” akjandara “gradu ally breaking out in thunder” (akja- “to thunder”); sakdara “ageing,”
gorel-1.p65
255
4/17/02, 8:57 AM
256
part four
sakdandara “gradually approaching old age” (sakda- “to get old,” “to age”); were “melting,” wendere “gradually approaching the melted condition” (we- “to melt”) (Zakharov, 1879:186). In the out-of-case form the imperfect participle is usually used as an attribute or a predicate: ài (ma. si) nene-me iàinji-ci (ma. isinji-ci) uthai you be.first-CONV come-CONV at.once àin-de (ma. sin-de) bu-re you(ài/àin- SG)-DAT give-PART “If you come first I shall give (something) back to you straight away;” tua-ci (ma. tuwa-ci) uhuri (ma. uhuri) fanga (ma. fangga) erdemu look-CONV every magic capability taci-re bithe bi-hebi learn-PART book be-PAST “When (one) had a look (then discovered) that those books were for learning magic” (SK: 65). The suffix for the perfect participle is -ha/-he/-ho, and for a small number of verbal stems it is -ka/-ke/-ko: alaha (ala- “to tell”), genehe (gene- “to go”), jodoho (jodo- “to weave”); badaraka (badara- “to expand”), duleke (dule- “to pass,” “to go by”), soroho, soroko (soro- “to turn yel low,” “to avoid as taboo”). According to Zakharov, in Manchu there are 185 verbs ending in -ka/-ke/-ko (87 verbs end in -ka, 81 verbs in -ke and 17 verbs end in -ko) (Zakharov, 1879:174). Even fewer num bers of verbs end in -ngka/-ngke/-ngko: bangka (ba- “to be lazy”), cangka (ca- “to stretch,” “to pitch a tent or yurt”), jengke (je- “to bear,” “to put up with,” “to tolerate”), jongko (jo- “to bring to mind,” “to re call,” “to mention”), jungka (ju- “to clench the teeth”), juwangka (juwa“to open the mouth”), fungke (fu- “to become numb”), sungke (su- “to freeze”), àangka (àa- “to cook dry,” “to drain dry”), àungke (àu- “to be thoroughly acquainted with,” “to know thoroughly”), wengke (we- “to melt”), yungke (yu- “to have a preference for,” “to be addicted to,” “to absorb (a dye),” “to soak in”). In the out-of-case form perfect participles usually act as attributes to nouns or predicates: emu mou (ma. moo) i dalba de emu sahal’an (ma. sahaliyan) one tree GEN near DAT one black n’alma (ma. niyalma) be sabu-ha man ACC see-PART “(Someone) saw a black man near a tree;”
gorel-1.p65
256
4/17/02, 8:57 AM
257
morphology
ere abala-me gehe-he gucu-sa ji-ci ai this hunt-CONV go-PART companion-PL come-CONV what seme ala-mbi tell-IMPF ai seme “why?,” “for what reason?” (NL:8); “What will (our) companions, who have gone hunting, tell (us) when they return?” (SK:64). 5.6.2. Participles in the Forms of Oblique Cases The Manchu participles may be followed by markers of the oblique cases when they function as predicates of clauses or objects and adverbial modifiers, either independently or together with depen dent words. Participles are often used in the dative. The imperfect participles followed by the dative have the meaning of an action which takes place during the time another action is performed or is a result of the completion of that action. Here are some examples: baita be deribu-re de, deribun be bodo-mbi matter ACC begin-PART DAT beginning ACC think-IMPF “When (someone) begins the matter (he) thinks about (its) beginning (how to begin this matter)” (PASH1:26). mama de ala-ra de, mama hendu-me … old.woman DAT tell-PART DAT old.woman say-CONV “When (he) has told to the old woman this old woman says … ;” jabu-re de jogån (ma. jugån) de ume jilgan tuci-re go-PART DAT road DAT do.not voice emerge-PART “Going along the road, (you) do not speak” (SK:66). The perfect participles in the dative usually denote actions after which another action starts: ajige ju-se damu sefu i jilgan be donji-ha little little.child/son(jui)-PL only teacher GEN voice ACC hear-PART de fayangga gemu tuhe-mbi se-cina DAT courage all collapse-IMPF say-OPT “Only boys heard the teacher’s voice all (of them) lost their cour age” (ZAKH:181); muse ere wakàam-be geli wa-ha de muke iningdari we this frog(wakàan)-ACC also kill-PART DAT water every.day
gorel-1.p65
257
4/17/02, 8:57 AM
258
part four
(ma. inenggidari) lakcarakå eye-mbi uninterruptedly flow-IMPF “After we kill these frogs water will flow uninterruptedly” (SK:68). The accusative is the second most commonly used marker participles may be followed by. Participles in the accusative usually serve as direct objects which have the meaning of an action or a person who is characterized according to the action he performs: in-i mute-re be ala-habi he(i/in-)-GEN be.able-PART ACC say-PAST “(They) said about his ability (to do smth.)” (ZAKH:211); ai turgun bisi-re be sa-rkå what reason be-PART ACC know-PART(NEG)
“I am unaware of any reason” (PASH:58);
muda-me jide-re be jogån (ma. jugån) de aca-ha
return-CONV go-PART ACC road DAT meet-PART
mangi (ma. manggi)
after
“After that (he) met (people who) returning on the road;”
ere taci-re be ahån de ulhibu-he this learn-PART ACC elder.brother DAT inform-PART akå there.is.not “(He) did not inform his elder brother about his learning “ (SK:66). Participles in other oblique cases are rarely found. They may be used with marker for the ablative and genitive: àun tuci-re ci jabu-me deribu-he the sun rise-PART ABL go-CONV begin-PART “After the sun rose (they) started;” muse de buce-re ci gåa (ma. gåwa) we (INCL) DAT die-PART ABL another arha (ma. arga) akå way there.is.not “There is no other way (to save) us from death” (SK:66). 5.6.3. Declension of Clauses with Participial Predicates Declension of participles and participial constructions quite often de velops into declension of sentences (clauses). As mentioned above, participles in the out-of-case form may be
gorel-1.p65
258
4/17/02, 8:57 AM
259
morphology
attributes or predicates in a sentence. Participles in the form of an oblique case may serve as objects or adverbial modifiers. Participles in the nominative may be used as subjects with which predicates that have semantics of evaluation are correlated. Secondary parts of the sentence expressed by the participle may be extended, and often form participial constructions which are governed by principal predicates or another part of a sentence. In this case declension of participles does not differ from noun declension since it expresses subordinate relationships within a sentence. The appearance of case markers on the boundary between two clauses is a more complicated situation in which case markers are formal devices used to subordinate a dependent clause to a principal one. Declension of participial constructions and declension of clauses with the participial predicate cannot be distinguished from a mor phological point of view, but they can be easily distinguished syn tactically. In a participial construction the case marker directly follows the participle, and from a syntactic point of view it belongs to the par ticipial construction which serves as a secondary part of a sentence and occurs between a subject and a predicate. In a clause, the case marker is placed between the dependent clause with the participial predicate and the principal clause and connects them with each other. If the case marker is removed, a complex sentence will be disinte grated into two independent parts. In Manchu, the complex (or compound) sentence presumably de veloped in two directions. Firstly, a complex sentence may be formed by extending the participial construction within a sentence. Secondly, a complex sentence may be organized by combining two simple sentences. In the Manchu language several types of complex sentences, containing certain case markers within their structure, exist. The presence of the participial predicate in a clause is evidence that it is the case marker which serves as a device which connects a depen dent and a principal clause and it is not the conjunction homony mous with it. The subordination of dependent clauses with the participial predi cate to principal clauses is carried out mostly by two case markers, namely by forms of the dative and the accusative. The marker of the dative is used after the participle in clauses that have temporal semantics. This marker is seldom used with the participle in object
gorel-1.p65
259
4/17/02, 8:57 AM
260
part four
clauses. The marker of the accusative is found with the participle in object clauses. Here are some examples: adaki boo-i anggasi hehe joodang be yarkiya-ra neighbor house-GEN widow woman Zhoodang ACC entice-PART de joodang cira be tob obu-fi halbu-rakå DAT Zhoodang face ACC right make-CONV go-PART(NEG) “When widow from the house next door enticed Zhaoodang, keep ing the honour he did not go (to her home)” (PASH1:262); boli be daha-ra emu niyalma morin i yarfun be hålha-ra Boli ACC follow one man horse GEN tether ACC steal-PART be temujin i deo belkutei sa-habi ACC Temuzhin GEN younger. brother Belkutei know-PAST “Belkutei, the Temuzhin’s younger brother, knew that a man fol lowing Boli (from the Boli’s suite) stole a tether” (PASH:45); àin-i (ma. sin-i) sain ara-ha de karula-me you(ài/àin- SG)-GEN good make-PART DAT repay-CONV ài (ma. si) ai ara se-ci ai ara-ki you(SG) what make say-CONV what make-OPT “Repaying you for the good you have done I shall make everything that you will say to do;” muduri ishunde beye beye haira-ra be dragon one.another self self love-PART ACC sa-habi understand-PAST
“The dragon understood how (they) love tenderly each other” (SK);
ere ejen’i sargan inengdari (ma. inenggidari) ere bou (ma. boo)
this khan(GEN) wife every.day this house
ici gene-re be tere daifu inengdari sabu-mbi toward go-PART ACC that doctor every.day see-IMPF “That doctor sees every day that this khan’s wife goes every day to this house” (SK). 5.6.4. Negative Participle Forms The negative participle is formed by the particle akå which also serves for the negation of the presence of an object or its quality. How ever, there is formal difference between negative forms of participles and nouns. In the structure of the negative participle the concre tion between the participial suffix and the negative particle occurred whereas with nouns the negative particle akå is used as a separate
gorel-1.p65
260
4/17/02, 8:57 AM
261
morphology
word. Presumably, in the past the negative particle was used with the participle as a separate word too. In Radlov’s records of Sibe, the negative particle is often written separately from the participle, for example: bure akå “(someone) does not give (smth.),” ulhibuke akå “(someone) did not give to know.” It can be conjectured that the negative particle akå preserves its lexical independence when it logi cally emphasized (Lebedeva & Gorelova, 1994:68). Usually the particle akå is not governed by the law of vowel harmony. However, in classical Manchu in most cases the particle akå is fully fused with the participle into one word: arahakå < araha + akå (ara- “to do,” “to make”); generakå < genere + akå, genehakå < genehe + akå (gene- “to go”); henduhekå < henduhe + akå (hendu- “to say,” “to speak”); ohakå < oho + akå (o- “to become”); sarakå < sara + akå (sa- “to know”), etc. Some perfect participles terminating in -ka/-ke/-ko change the con sonant k into h in their negative form: jongko + akå > jongkakå/jonghakå (jo- “to recall,” “to mention”); soyoko + akå > soyohakå (soyo- “to shrink,” “to wrinkle”); talmaka + akå > talmahakå (talma- “to be foggy”); tucike + akå > tucihekå (tuci- “to appear,” “to emerge,” “to originate from”). In complex participial forms which are formed by perfect participles derived from autonomous verbal stems and by those derived from auxiliary verbal stems, the negative particle akå is fused with the first component: alahakå bihe, genehakå bihe. Instead of an affirmative utterance the Manchus widely used a so-called double negation where the first negative particle is writ ten together with the participles and the second one occurs sepa rately after them: arahakå akå, genehakå akå, sarakå akå (Zakharov, 1879:213-4). The negative particle unde “not yet” is normally used after the im perfect participle even when it denotes an action performed in the past: jidere unde “(someone) has not come yet.” Here is a sentence which includes the negative particle “unde:” sin-i gisun angga ci tuci-bu-re unde you(SG)-GEN word mouth ABL appear-PASS-PART not.yet bisi-re-de, i je inu be-PART-DAT he yes yes be aifini belhe-me jabdu-ha ACC already prepare-CONV complete-PART “Although no word has been yet said by you, he has already pre pared the answer: yes, exactly!” (ZAKH:215).
gorel-1.p65
261
4/17/02, 8:57 AM
262
part four
Only the participle is characterized by the opposition between posi tive and negative forms. Neither verbs nor converbs have negative forms. Therefore, in cases when the negative form of the verb or the converb is required, the negative participle is used. Along with this pattern of verbal forms formation, Norman lists another one, discovered by him in the Sibe dialect, according to which the negative suffix -qu (< aqu) is attached directly to the ver bal stem: bii saqu “I do not know.” The negative form of the exis tential verb bii “there is/there are” is aqu. According to him, to obtain various modalities of the negative form, the appropriate form of the verb o- “to become” in its auxiliary function is used: ar�qu oki “Let’s not do it;” ar�qu oci > ar�quci “if you don’t do it” (Norman, 1974:172). Negative participles may serve as any part of a sentence, how ever, they are mostly used as a predicate of a sentence since they may replace the forms of the verb proper. Here are some examples: cananggi bu-mbi se-ci bu-hekå, sikse previously give-IMPF say-CONV give-PART(NEG) yesterday bu-mbi se-ci geli bu-hekå give-IMPF say-CONV again give-PART(NEG)
“Previously (you) said that (you) would give (smth.), but didn’t give;
yesterday (you) also said that (you) would give (smth.), but again didn’t
give” (PASH2:131);
yamji cimari erge-re be bai-rakå evening morning rest-PART ACC seek-PART(NEG)
“(Myself) didn’t seek rest (peace) from morning till evening”
(PASH2:328);
ajigen-ci ere i adali fergecuke (ma. ferguwecuke) gaska
childhood-ABL this GEN like amazing bird
(ma. gasha) be àue (ma. àuwe) sabu-rakå
ACC not.at.all see-PART(NEG) “From the childhood (I) have not seen such an amazing bird, not at all;” ài (ma. si) ere erin de han’i fuyan o-ho-be you this time DAT khan(GEN) wife become-PART-ACC
dahame mim-be cihala-mbi cihala-rakå
because I(bi/min-)-ACC love-IMPF love-PART(NEG)
“Because you became the khan’s wife do (you) love me or not?”
(SK:69).
gorel-1.p65
262
4/17/02, 8:57 AM
263
morphology
The adverbialization of negative participles frequently occurs due to the fact that negative participles often function as adverbs in a sentence: muke inengdari (ma. inenggidari) lakcarakå eye-mbi water every.day uninterruptedly flow-IMPF lakcarakå “uninterruptedly” < lakca-ra akå (lakca- “to break off”) “Water every day will flow uninterruptedly” (SK). The fact that verbs and converbs have no negative forms is addi tional evidence for the late formation of verbs proper and converbs as compared with the participle. 5.6.5. Substantive Forms of Participles and the Process of Nominalization of Participles In the Manchu language the substantive-possessive suffix -ngge which may be attached to both participles, perfect and imperfect (forms in -ha/-he/-ho and -ra/-re/-ro respectively) and to the negative forms of these participles. This suffix does not change its vowel according to the law of vowel harmony. Normally it is written together with participles. As mentioned above, from the diachronic point of view, the suf fix -ngge was related to the meaning of possession (see Section 2.4.1). Following a single participle or a subordinate predicative construc tion with a participle as a head, the suffix -ngge permits both of them to play the part of a subject or an object in a sentence. I shall use the term “nominalization” to designate the process by which a word of any grammatical class may act (and the result of this acting) in the noun position (see also Section 2.4.1). From the synchronic point of view, the suffix -ngge can rightfully be called a nominalizer. The nominalization of participles (and the subsequent process which is substantivization, i.e. the shift of a word to the grammati cal class of substantives) has developed in several directions. Parti ciples followed by the suffix -ngge may denote: 1) an abstract notion about an action and can be translated as abstract substantives: aliburangge “presentation of a document to a superior;” bairengge “an application submitted to a superior” or 2) an object which, in its turn, may be of material or immaterial nature: minde buhengge “what was given to me,” “present,” “what you have done,” “a deed,” 3) a subject: “one who says, thinks, etc.,” “those who say, think, etc.” Not only affirmative participial forms in -ngge but also their negative
gorel-1.p65
263
4/17/02, 8:57 AM
264
part four
forms may be characterized by these meanings: akdarakångge “distrust,” “one who does not trust;” saburakångge “what is not seen,” “never seen before.” In certain cases, participles followed by the suffix -ngge may denote both an abstract action and a person who is characterized according to the action he performs. The difference between these meanings can be understood only from the syntactic context: saisa be huwekiyebu-re-ngge wise.man/scholar ACC incite-PART-NR “inciting wise men/scholars” or “one who incites wise men/scholars;” niyamniyam-ra-ngge gabta-ra-ngge shoot (arrows).from.horseback-PART-NR shoot.an.arrow-PART-NR gemu sain all good “Mounted and unmounted shooting are both good;” “One who shoots from horseback and unmountedly is good in all respects” (ZAKH:76). Here are some sentences where nominalizations display their different meanings: wesi-re forgoào-ro-ngge umesi hådun promote-PART transfer-PART-NR very quick “Promotions and transferences are very quick” (ZAKH:76); bou (ma. boo) de muda-ci tuci-re-nge (ma. tuci-re-ngge) house DAT return-CONV leave-PART-NR manga (ma. mangga) o-mbi hard be-IMPF “It will be hard to leave (again) after returning home;” yabu-ha-nge (ma. yabu-ha-ngge) yarg’an’i (ma. yargiyani) tondo do-PART-NR indeed honest akå there.is.not
“Our action (what we did) is not honest indeed;”
ere be muke se-me omi-ha-nge (ma. omi-ha-ngge)
this ACC water say(AUX)-CONV drink-PART-NR
gemu araki bi-hebi everything wine be-PAST “Everything that (he) drank thinking it was water was wine” (SK:70).
gorel-1.p65
264
4/17/02, 8:57 AM
265
morphology
A participle construction with a participle as a head, may contain its own subject expressed by a noun in the genitive case: age i jombu-re-ngge aisin go i gese elder.brother GEN suggest-PART-NR gold jasper GEN like gisun kai words COP “The elder brother’s words are comparable to gold and jasper (What the elder brother said are the words like gold and jasper)” (Pang, 1986:94); han’i nime-re-nge (ma. nime-re-ngge) majige sain khan-GEN ill-PART-NR a.little good o-ho become-PART “The khan’s illness became a little better” (SK:70). The absence of any case markers after participles in the form in -ngge denotes the nominative which normally marks a subject. The marker of the accusative, added to participles, marks an object: geren hafa-sa cooha nirgen ere sargan’i jide-re-nge many official-PL soldier people this woman(GEN) come-PART-NR (ma. jide-re-ngge) be goroki-ci sabu-fi hendu-me … ACC distant.place-ABL see-CONV say-CONV “Many officials and soldiers saw from afar that this woman was coming (this woman’s coming) and said …” (SK:71). Direct speech, quotations, hearsay are often introduced with parti ciples in -ngge. In this function participles derived from vebs of speech are mostly used: ala-ra-ngge (ala- “to say,” “to tell”), hendu-re-ngge (hendu“speak,” “to talk”), se-re-ngge (se- “to say”), fonji-re-ngge (fonji- “to ask”), wesimbu-re-ngge (wesimbu- “to report”). Participles in the form in -ngge are placed before direct speech: yuan fai i sargan jui hendu-re-nge (ma. hendu-re-ngge) proper name GEN female child speak-PART-NR tere ahån deo hadan niyalma gemu håsun that elder.brother younger.brother seven man all power amba niyalma big man “This is what Yuan Fai’s daughter says: all those seven brothers are powerful and big men;”
gorel-1.p65
265
4/17/02, 8:57 AM
266
part four
ere fuyanå (ma. fiyanggå) sargan jui dobori ere hotun’i (ma. hoton) this youngest female child at.night this walled.city(GEN) emu sagda (ma. sakda) mama jakade gene-fi one old old.woman to go-CONV songo-mo (ma. songgo-me) ala-ra-nge (ma. ala-ra-ngge) … cry-CONV tell-PART-NR “This is what the youngest daughter who came at night to the city’s old woman tells crying” (SK:71). In the Sibe dialect investigated by Norman, the suffix -nge (-ngge in Manchu) is used as a perfect finite form. This function is common for sentences which express time or place semantics: juu (ma. juwe) ani (ma. aniya) t� -x� -�� bii b� ijing-de I Peking-DAT two year live-(PART-nge) =PERF “I lived in Peking for two years” (Norman, 1974:173). In the Manchu language the syntactic element ba is used to sub stantivize participles. This substantivizer is used with perfect and imperfect participles, their affirmative and negative forms: bodoro ba “what is planned,” “plan;” gåniha ba “what is thought,” “thought;” sere ba “what is said,” “speech.” Combinations of participles with the element ba have occasional, not steady character. Here are some examples: alibu-ha ba bi-he present (a document to superior)-PART SBSTR be-PART “There was a report to authorities;” afabu-ha ba kemuni bi order-PART SBSTR still COP “There is an order (from authorities)” (ZAKH:76); àin-de (ma. sin-de) ai gåni-ha ba bi you(ài/àin- SG)-DAT what think-PART SBSTR COP gåni-ha ba “(you) have thought,” “your thoughts;” “What thoughts have you got?;” te uàin (ma. usin) weile-me gene-ci ai now field cultivate-CONV go-CONV what ojo-rakå se-re ba bi what became-PART(NEG) say-PART SBSTR COP se-re ba “what is said,” “talk,” “speech;” “How can one talk about the impossibility of going to cultivate a field?” (SK:71).
gorel-1.p65
266
4/17/02, 8:57 AM
267
morphology 5.7. Converbs
Manchu converbs can be categorized as a separate subclass of ver bal forms. It should be particularly pointed out that converbs, as a special class of verbal forms, which function as means of subordi nation of one verb to another, are characterized by the existence of agent/subject valencies. In other words, Manchu converbs may be oriented towards the subject (agent) of the dominant verb, but they may also have their own grammatical subject referentially indepen dent from the subject of the principal action. This subject, howev er, due to certain specific features of Manchu grammar, cannot be reflected in the converb form itself by the possessive-predicative suffixes (as is possible in other Tungusic languages). It may be ex pressed by either an independent word form or it is not necessarily expressed explicitly at all. In the latter case the subject is reconstruct ed from the syntactic context. The major feature distinguishing the converb from the verb proper is that converbs may not serve as a predicate of a simple sentence. They may function as adverbs or predicates of adverbial clauses. They may also be used in analytical verbal constructions, which have different temporal, aspectual or modal semantics. In Manchu there are two widely used converbs, i.e. the imperfect and the perfect converbs. 5.7.1. Imperfect Converbs Being one of the most commonly used forms in Manchu, the imperfect converb realizes a wide and diverse range of functions. The main meaning of the imperfect converb the suffix for which is -me, is that of simultaneity of both dependent and principal actions. In other words, the converb of this type, as a dependent (non-finite) form, denotes relative tense which, as known, is not related to the moment of speaking but to the tense of the finite form. The temporal characteristics of dependent action expressed by the imperfect converb are transmitted to this action from the finite form and from the tense characteristics it possesses. Thus, simultaneity may occur in the past, present, and future. Empirical material connected with the functioning of the imperfect converb proves that the hypothesis of broader semantics of simple forms is correct. Thus, the simpler the form is structurally, the more developed system of functions it has (Shamina, 1985:5). On the basis of its main meaning, that of simultaneity of both
gorel-2.p65
267
4/17/02, 10:59 AM
268
part four
subordinate and principal actions, the imperfect converb develops a wide range of other adverbial meanings, firstly meanings of manner in which another action is performed; condition, cause, and purpose. Here are some examples: bithe de foro-me, hergen be tolo-me, yasa book DAT turn-CONV letter ACC count-CONV eye tuwa-me observe-CONV gala jori-me urebu-me håla-mbi hand point-CONV rehearse-CONV read-IMRF “Turning toward the book, counting written characters, observing (closely) with eyes, pointing with his hand (finger), (someone) reads (the text) rehearsing” (ZAKH:190); muse sibiya makta-me dasame dende-ki we(INCL) lots throw-CONV again divide-OPT “Casting lots we shall divide (smth.) again” (ZAKH:191); taidzung genggiyen s�u håwangdi, banitai umesi Taizong enlightened educated emperor naturally very enduringge ofi, abka-i forgon be badarambu-me holy as Heaven-GEN favour ACC propagate-CONV nei-he, gånin be sira-me, erdemu be disclose-PART idea ACC follow.up-CONV valour ACC fisembu-me, s�u i dasan be ambara-me increase-CONV culture GEN rule ACC spread-CONV selgiye-he promulgate-PART
s�u genggiyen “cultured and enlightened” (NL:263);
“As Taizong, enlightened emperor, was naturally very wise, [he]
propagating disclosed the Heaven’s favours, following up [the same]
ideas, increasing valours, spreading [widely], promulgated his
enlightened rule” (PASH1:321; literal translation);
“As Taizong, enlightened emperor, was naturally very wise, he spread
the word of Heaven’s favours, and building on these ideas with
increasing courage extended his enlightened rule” (literary translation);
tere-ci enen akå jalin facihiyas�a-me, yuan wai eigen that-ABL heir there.is.not since grieve-CONV yuanwai husband sargan damu sain be yabu-me, juktehen be niyece-me wife only good ACC do-CONV shrine ACC nourish-CONV
gorel-2.p65
268
4/17/02, 10:59 AM
269
morphology
weile-me, fucihi de kesi bai-me construct-CONV Buddha DAT happiness seek-CONV hengkis�e-me kowtow.repeatedly-CONV enduri de jalbari-me, yan hiyan be jafa-fi, ba God DAT pray-CONV ounce(?) incense ACC take-CONV place ba-de hiyan dabu-me, geli yadahån urse aisila-me, place-DAT incense burn-CONV also poor people help-CONV umudu be weihiye-me anggasi be aitubu-me sain be orphan ACC support-CONV widow ACC revive-CONV good ACC yabu-fi iletule-re jakade, dergi abka gosi-fi do-CONV expose-PART since High Heaven take.mercy-CONV yuanwai < chin. “official;” “As there was no heir, grief-stricken yuanwai and his wife, by doing good, helping to build a shrine, praying to Buddha, collecting incense candles, burning incense, begging happiness for kind of spirits, helping poor people as well, supporting orphans, and reviving widows, became famous for their good works and the High Heaven took mercy on them” (NSB:111); tumen baita icihiya-ha s�olo de, damu bithe be myriads deed manage-PART free.time DAT only book ACC håla-me, giyan be kimci-me gånin be read-CONV principle ACC scrutinize-CONV mind ACC akumbu-ha strain-PART “In the periods when free of managing myriads of deeds, (I) only reading books and scrutinizing the truth, strained (my) mind” (PASH1:325); ere unin deberen banji-rakå o-ci bi this cow baby-calf give.birth-PART(NEG be(AUX)-CONV I omihulu-me (ma. omiholo-) lala-me (ma. lali-) suffer.hunger-CONV be.weakened.from.hungry-CONV kangka-me buce-mbi be.thirsty-CONV die-IMPF “If this cow does not give birth to a baby-calf, I will die of hunger and thirst (suffering hunger, being thirsty)” (SK). The imperfect converb is commonly used with verbs of motion, phase and modal verbs, and their functions are analogous to those of the Russian infinitive. Here are a few examples of the use of the imperfect
gorel-2.p65
269
4/17/02, 10:59 AM
270
part four
converb followed by verbs that have semantics of motion. In the following constructions imperfect verbs denote the purpose of an action: morin be dali-me boo-de ji-he horse ACC hide-CONV house-DAT come-PART “To hide the horses, [he] went home” (PASH1:173); ere adaki boo-i anggasi hehe mim-be yarkiya-me this lovely house-GEN widow woman I(bi/min-)-ACC entice-CONV ji-he bi-he come-PART be-PART “This lovely woman-next-door came to entice me” (PASH2:262; CT); sa-rkå ojo-ro jakade tuttu urgun know-PART(NEG) be(AUX)-PART as so joy ara-me gene-hekå make-CONV go-PART(NEG)
ojoro jakade “as,” “since;” urgun ara- “to congratulate,” “to wish well;”
“As (I) did not know (about your arrival), so (I) did not go to
congratulate you” (PASH:60);
ere mama muke gaji-me gene-fi muke be this old.woman water fetch-CONV go-CONV water ACC hunio de tebu-me funtoko deri emu dube i basket DAT scoop-CONV hole SEP one side GEN eye-mbi run-IMPF “This old. woman went to fetch some water, and scooped it up with a (water) bucket, and the water is running through a hole at one side” (SK). Here are some examples of the use of the imperfect converb followed by phase verbs: amba edun da-me deribu-he high wind blow-CONV begin-PART “High wind began blowing” (ORL:176); usin tari-me boo ara-me deribu-he field cultivate-CONV house build-CONV begin-PART “(He) began to cultivate the field and to build a house” (PASH1:91); deo gosiholo-me naka-rakå younger.brother grieve-CONV stop-PART(NEG) “(My) younger brother did not stop grieving” (PASH1:266).
gorel-2.p65
270
4/17/02, 10:59 AM
271
morphology
The examples given below include imperfect converbs followed by modal verbs: bi nure omi-me bahana-rakå I wine drink-CONV be.able-PART(NEG)
“I am not able to drink wine” (PASH:53);
muke gaji-me mute-hakå min-de
water bring-CONV be.able-PART(NEG) I(bi/min-)-DAT
bu-fi ulubu-rakå (ma. ulebu-)
give-CONV feed-PART(NEG)
“As you were not able to bring water, I shall not give you anything
to eat;”
ere buce-rakå o-ci min’i ba’i (ma. biyai)
this die-PART be(AUX)-CONV I(bi/min-)-GEN moon-GEN
elden saurin (ma. soorin) àira-me (ma. sira-me) mute-rakå
light throne inherit-CONV be.able-PART(NEG)
“If he does not die, my Light of the Moon (proper name – L.G.)
won’t be able to inherit the throne” (SK).
In combination with verbs of broad semantics, like tuci- “to appear,”
“to go;” gene- “to go,” “to set off;” ji- “to come,” “to arrive;” gai-
“to take;” wa- “to kill” and others, imperfect converbs derived from
verbs of concrete semantics are used to specify actions of broad
semantics.
Here are the combinations of the imperfect converb with various forms of the verb tuci-: uka-me tuci-he ( running away went out > ran away); ana-me tuci-bu-he (pushing let smb. appear > pushed out). The following combinations include the imperfect converb with the verb gai-: uda-me gai-ha (buying took > bought); hålha-me gai-ha (robbing took > robbed). The following combinations consist of the imperfect converb with the verb gene-: amca-me gene-he (running down went > reached); dosi me gene-he (entering went > entered). Here are the combinations of the imperfect converb with various forms of the verb ji-: suju-me jiu (running come > run [here]); muda me ji-he (returning came > returned). The following combinations consist of the imperfect converb and the verb wa-: gapta-me wa-ki (shooting kill > shoot), saci-me wa (killing with a sabre kill > kill with a sabre). Here are some examples:
gorel-2.p65
271
4/17/02, 10:59 AM
272
part four
emu gucihe o-fi deye-me yabu-ha one pigeon become-CONV fly-CONV go-PART “Having turned into a pigeon, (one) flew off (flying off went);” goidahakå abala-me gene-he ilan n’alma (ma. niyalma) in.a.short.while hunt-CONV go-PART three man muda-me ji-he mangi (ma. manggi) ere ilan niyalma return-CONV come-PART after this three man fonji-me ask-CONV “After in a short while three people who had gone hunting returned (returning came back), these three men ask” (SK); Combination of the imperfect converbs of synonymous verbs is more frequently used to express a higher level of action intensity: gemu songo-me (ma. songgo-me) fa-me men’i men’i everyone cry-CONV sob-CONV each/every bou (ma. boo) i baru muda-ha house GEN to come-PART “Everyone shedding bitter tears (crying, sobbing) went home, each to their own place;” bou (ma. boo) i doài-ha (ma. dosi-ha) be tere taitai house GEN fly-PART ACC that mistress sabu-fi umuài (ma. umesi) urgunje-me bayala-me see-CONV very be.glad-CONV be.happy-CONV “The mistress having seen that (the bird) flew inside the house was very pleased” (SK). The imperfect converb is frequently used in combination with different forms of the verb bi- “to be” in its auxiliary function. The functions of such newly created analytical forms require more detailed study, however, we know that they have aspectual-temporal meaning in the first place, as in: ala-me bi-me “telling;” banji-me bi-he “having lived long;” gene-me bi-he “having gone.” The form of the imperfect converb may be fused with grammatical forms of the verb bi- in its auxiliary function, for example: alambime (< ala-me bi-me) “telling;” alambihe (< ala-me bi-he) “having said;” bahambihe (< baha-me bi-he) “having found;” wambihe (< wa-me bi-he) “having killed;” gasambihe (< gasa-me bi-he) “having grieved.” In connection with such forms it should be mentioned that some of them are considered independent
gorel-2.p65
272
4/17/02, 10:59 AM
273
morphology
converbs, for example, the form in -mbime (Zakharov, 1879:191). The form of the imperfect converb derived from verbs of locution (verba dicendi) such as ala- “to tell,” “to report;” gisure- “to speak,” “to talk;” hendu- “to say,” “to speak” is normally used to introduce direct speech, placed in postposition to these verbs. Here are some examples: Belkutei daha-ra urse afa-ki se-re-de, PN follow-PART people attack-OPT say(AUX)-PART-DAT belkutei bejige-me hendu-me suwe ainu PN persuade-CONV say-CONV you(PL) why ehe ara-ra be buye-mbi, min-i feye evil make-PART ACC like-IMPF I(bi/min-)-GEN=my wound ujele-hekå-bi naka se-ci be.serious-PART(NEG)-COP stop(IMP) say(AUX)-CONV Tv-ki + se- “to want, to desire to do smth.;” “When people who followed Belkutei wanted to attack him, Belkutei said: “Why do you like to make evil, my wound is not serious, stop (it)!” (PASH1:10); tere Sidi Kur folho i dorgi de bi-fi hendu-me, that proper name sack GEN inside DAT be-CONV say-CONV akå (ma. ahån) ài (ma. si) àim-be (ma. sim-be) meihere-me brother you you(ài/àin-SG)-ACC shoulder-CONV yabu-ru (ma. yabure) de sada-rakå na walk-PART DAT be.tied-PART(NEG) INT “That Sidi Kur, being inside a sack, says: Brother, you have shouldered me and are walking, aren’t you tired?” tere han gisure-me tere sargan jui hairakan that khan say-CONV that female child regrettable “That khan says: (I) feel sorry for that girl!” (SK). The form of the imperfect converb derived from the verb of locution of broad semantics se- “to say,” “to speak” may be used to introduce subordinate predication which represents information on real events in a peculiar way so that the opposition between direct and indirect speech is neutralized. In this case the form seme is placed in postposition to the text containing the opposition “direct speech : indirect speech,” for example:
gorel-2.p65
273
4/17/02, 10:59 AM
274
part four
doose geren i baru suwe omi-me ele-he-o monk everybody GEN toward you(PL) drink-CONV be.enoughPART-INT se-me fonji-ha de, elhe se-me jabu-ha say-CONV ask-PART DAT well say-CONV answer-PART “[When] the monk asked everybody: “Have you had enough to drink?,” [they] answered: “Enough” (PASH:61). Constructions of this type form the basis for using the word form seme as an analogue of a conjunctive-analytical device (a specific analogue of expounding conjunctions such as “that,” “in order to,” “how”): enenggi erde sakda ahån be today early.in.the morning old.man elder.brother/sir ACC isinji-ha seme donji-fi cohotoi dorolo-me arrive-PART that hear-CONV especially greet-CONV ji-he come-PART
“Today early in the morning (I) heard that you, sir, arrived, and (I)
came especially to greet you” (PASH:57).
With negative forms of imperfect participles as well as with the form
in -rahå which is originally the negative form of the imperfect
participle (-rahå < -ra/-re/-ro + akå), the word seme forms a particular
meaning designating apprehension that a certain action might take
place (see also Section 5.8.3):
ama eme damu nime-rahå seme jobo-mbi father mother only ill-PART(NEG) that worry-IMPF “My only worry is that (my) father and mother might get ill” (ZAKH:210); ere jalan de bi àim-be (ma. sim-be) sabu-rakå this world DAT I you(ài/àin- SG)-ACC see-PART(NEG) seme gåni-ha bi-he that think-PART be-PART “I thought that I might never see you again in this world” (SK). The converbal form seme may develop conjunctive meaning on the basis of its primary meaning of the converb proper, also in another
gorel-2.p65
274
4/17/02, 10:59 AM
275
morphology
direction, resulting in a conjunctive device that have specific
conditional meaning (interim between real and unreal), and,
furthermore, concessive meaning. The following complex sentences
should be referred to the semantic space of unreality, and the word
form seme has the meaning “even if”:
gama-ha seme tusa akå
take-PART even.if use COP.NEG (there.is.not)
“Even if (I) took (smth.) it would be of no use” (NSB:129);
i gene-he seme, inu baita de tusa ara-ra ba he go-PART even.if also matter DAT use make-PART SBSTR akå COP.NEG. (there.is.not)
“Even if he went (there) it would be of no use” (ZAKH:207).
Another opportunity to analyze the word form seme within these
complex sentences is to consider it as being a part of the analytical
construction Tv-ha seme (see Part V, Section 5.6). Most likely that
seme, as fixed within constructions of this type, was in its way of the
transition to a syntactic word of the conjunctional type.
In Manchu, apart from the above-mentioned functions, seme frequently marks the topic (for details see Part V, Section 3): uttu o-ho-de, teni niyalma seme jalan de so be(AUX)-PART-DAT then man TOP world DAT banji-re de yertecun akå o-mbi live-PART DAT shame there.is.not become-IMPF “If so, then a man, while he lives in the world, will not be ashamed (shame will become non-existent)” (GAB:120). Finally, seme regularly participates in the formation of onomatopoetic words (for a more detailed description of this function see Section 7). The connection of these two functions of this form with its primary function of the converb proper is obvious. Therefore, one can distinctly trace the process of formation, on the basis of one word form seme and its primary function, of a whole complex of language units materially similar to and functionally connected with seme, but having different grammatical status and different functions. These units realize their meaning in a great number of grammatical structures, and this trend determines, to a considerable extent, the
gorel-2.p65
275
4/17/02, 10:59 AM
276
part four
specific feature of Manchu grammar—that of polyfunctionality of language units (Gorelova, 1987). 5.7.2. Perfect Converbs The suffix for the perfect converb is -fi: alafi (ala- “to tell,” “to report”); isinafi (isina- “to reach,” “to arrive”); jefi (je- “to eat”); obofi (obo- “to wash”); tefi (te- “to sit,” “to live”). It is the second most widely used converb in Manchu: si boo-de isina-fi, majige teye-fi, buda je-fi you house-DAT arrive-CONV a.little rest-CONV food eat-CONV dere obo-fi jai ji-ki face wash-CONV then come-OPT “After you having arrived home, rested a little, eaten a meal, washed (your) face, then come” (ZAKH:191). Some verbs have an irregular suffix for the perfect converb, namely -pi or -mpi: juwampi (juwa- “to open”); sampi (sa- “to stretch”); s�ampi (s�a- “to cook dry,” “to drain dry”); wempi (we- “to melt”). Here is an example: monggon sa-mpi angga juwa-mpi tuwa-mbi neck stretch-CONV mouth open-CONV observe-IMPF “I observe having stretched the neck and opened the mouth” (ZAKH:193). These verbs usually take irregular suffixes for the imperfect and perfect participles (-ndara and -ngka respectively). The suffix -pi is sometimes used with verbs which have regular perfect participial forms, but in this case it denotes a certain tension of an action or a superlative degree of quality of being or a principal action: colgoro-pi geren niyalma ci ujula-me tuci-kebi excel-CONV all people ABL head-CONV appear-PAST “Having excelled, he was above all people” (ZAKH:193). The perfect converb denotes an action which is completed before another action is undertaken. The dependence between the subordinate action, expressed by the perfect converb, and the principal action, mostly has temporal meaning, for example:
gorel-2.p65
276
4/17/02, 10:59 AM
277
morphology
bithe håla-ra niyalma bolgo boo-i dolo eici nahan book read-PART man clean house-GEN inside or oven-bed eici nikeku mulan de te-fi dere-i ninggude or chair DAT sit-CONV table-GEN on.the.top.of fi, yuwan, behe hoosan be sinda-fi writing.brush ink-pot Indian.ink paper ACC put-CONV bithe be sara-fi tuwa-me håla-mbi book ACC open-CONV look-CONV read-IMPF nikeku mulan “a chair” (NL:212); “A learned scholar, inside a clean room having sat on the oven-bed or the chair, put a writing brush, an ink-pot, Indian ink and a paper on the table, opened a book, (usually) reads” (ZAKH:192); tere nadan g’ahån (ma. giyahån) nadan n’alma (ma. niyalma) that seven hawk seven man o-fi àin-i (ma. sin-i) tere yirge (ma. ira) be become-CONV you(ài/àin- SG)-GEN that millet ACC bai-mbi look.for-IMPF “Those seven hawks having turned into people, would look for your millet;” ejen tulgide tuci-fi tua-ha (ma. tuwa-ha) khan outside go.out-CONV look-PART “Having gone outside, the khan looked at (smth.)” (SK:72). Derived from certain verbs in their auxiliary functions, the perfect converb may indicate the reason for performing an action. Such forms show a tendency to develop gradually into analytical markers of the conjunctive variety. Most Manchu conjunctions originated from such converbal forms (see also Sections specially dedicated to analytical converbal forms and conjunctions, 5.9 & 8.3 respectively): si tob sere niyalma o-fi, bi teni uttu tafula-ra you upright man be-CONV I then like.this advise-PART dabala MDL.PTL “I advised you (to do) like this because you are an upright man” (ZAKH:192). According to Zakharov, the verb aina- “to do what?,” “how? how is/are ... ?,” “what’s up?,” “why?” always requires the perfect converbal form from the dependent verb:
gorel-2.p65
277
4/25/02, 12:26 PM
278
part four
ere-be gisure-fi aina-mbi this-ACC speak-CONV do.what-IMPF (literally: “Having said this, do (you) do what?”); “Why do you speak this?” (ZAKH:193). i emgeri sa-fi goida-ha, si kemuni he already know-CONV last.for.a.long.time-PART you(SG) yet ala-rakå o-fi aina-mbi tell-PART(NEG) be(AUX)-CONV do.what-IMPF “He had already been known (that) for a long time, why haven’t you told (that) yet?” (ZAKH:192). 5.7.3. Conditional Converbs The suffix for the conditional converb is -ci: alaci “when/if one tells;” geneci “when/if one goes;” dosici “when/if one enters.” This form has mostly been considered by manchurologists as the form of the conditional mood (Zakharov, 1879:198). Mo�llendorff was of the view that this form constituted the conditional tense: araci “I should write,” “if I wrote,” “should I wrote” (Mo�llendorff, 1892:9). However, as this form may not function as a predicate of a simple sentence, it should not be analyzed as a mood marker. Being a non-finite form, it may serve only as a subordinate predicate in a conditional construction or a clause. Consequently, it is justifiable to include -ci with converbs. The conditional converb indicates that the subordinate action precedes the principal action in time. The additional shade of the conditional meaning is what distinguishes the conditional form from the perfect converb. Here are some examples: gene-ci uthai gene-mbi seme hendu, go-CONV then go-IMPF that say(IMP) gene-rakå o-ci, uthai gene-rakå seme go-PART(NEG) be(AUX)-CONV then go-PART(NEG) that hendu say(IMP)
“If (you are intending to) go, then say that (you) will go; if (you are
not intending to) go, then say that (you) will not go” (ZAKH:199;
QW);
muse uka-me tuci-ci sain o-mbi we run.away-CONV go.out-CONV good become-IMPF “If we run away it will be good;”
gorel-2.p65
278
4/17/02, 10:59 AM
279
morphology
ài (ma. si) ere ba-de banji-me cihala-ci men’i you(SG) this place-DAT live-CONV want-CONV we(EXCL)-GEN ere abka-i ba-de sasa banji-ki this heaven-GEN place-DAT together live-OPT “If you want to live in this place, (then) live in heaven together with me” (SK:74). The verbs aca- “to meet,” “to get together,” “to combine,” “to correspond to,” etc. and o- “to become,” “to be,” being partly desemantized, always require the conditional converb before them. The whole construction has the following meanings: “should,” “ought” or “must” (Tv-CONV aca- “should do smth”): bithe-i hafan o-ci, irgen be book-GEN official be-CONV the.common.people ACC gosi-me, giyan be bai-me icihiya-ci cherish-CONV order ACC seek-CONV manage-CONV aca-mbi AUX-IMPF
“If (you ) are a civil official, (then) cherishing the common people,
seeking after order, (you) should manage affairs;”
cooha-i hafan o-ci, cooha-i urse be military-GEN official be-CONV military-GEN people ACC tacihiya-me urebu-me, meni meni afa-ha train-CONV practice-CONV every fight-PART ba-be tuwakiya-ci aca-mbi place-ACC guard-CONV AUX-IMPF
meni meni “every,” “each” (NL:197);
“If (you) are a military official, (then) training military people, (you)
should guard every place where (military people) are fighting”
(ZAKH:205).
From a historical point of view, it is likely that the suffix of the
conditional converb is connected with the marker of the ablative
ci. These elements are not only materially identical but they have certain semantic similarity since the temporal meaning of the ablative case is sometimes very close to that of the conditional converb. Compare, for example, the use of the participle in the form of the ablative case with the use of the conditional converb:
gorel-2.p65
279
4/17/02, 10:59 AM
280
part four
s�un tuci-re ci yabu-me deribu-he sun rise-PART ABL go-CONV start-PART “When the sun rose (they) started;” bou (ma. boo) de ji-ci umai jaka hacin house DAT come-CONV not.at.all thing various n’alma (ma. niyalma) s�ue (ma. s�uwe) akå people not.at.all there.is.not “When one comes back home, there are no people and things, not at all” (SK:75). In Manchu, verbal stems cannot be declined, and this is the only obstacle to bringing together the ablative and the conditional form. Among all verbal words only participles can be declined. The question of whether or not verbal words denoting secondary parts of the sentence could be materially identical and declined in the past, has not yet been investigated. 5.7.4. Concessive Converbs The concessive converb is formed by the suffix -cibe. Presumably, the suffix -cibe consists of two elements, -ci and be, and the first element -ci has conditional meaning. Another interpretation of the origin of this form is also possible: the conditional stem in -ci is followed by the component be which have concessive meaning. In Turkic languages the concessive forms are formed in a similar fashion. A hypothesis about connection between the suffixes for conditional and concessive meanings has semantic justification since concessive semantics naturally follows from the conditional meaning. The following sentences contain concessive converbs: fulun caliyan bi-cibe, inenggidari baitala-rakå salary just.enough exist-CONV every.day use-PART(NEG) de isi-rakå DAT suffice-PART(NEG)
“There is a salary, but it is not sufficient for day-to-day necessities”
(ZAKH:206);
tere jue (ma. juwe) n’alma (ma. niyalma) ere ba-de bi-cibe
that two man this place-DAT be-CONV
s�ue (ma. s�uwe) sabu-rakå o-ho
not at.all see-PART(NEG) be(AUX)-PART
gorel-2.p65
280
4/17/02, 10:59 AM
281
morphology
“Although those two men were at this place (here) (they) were not seen at all;” nadan aniya tacibu-cibe jinkin’i erdemu be tacibu-rakå seven year teach-CONV in.fact skill ACC teach-PART(NEG) “Although studied for seven years, in fact he did not learn any skills” (SK:75). 5.7.5. Converbal Forms in -hai/-hei/-hoi The converbal form in -hai/-hei/-hoi or -kai/-kei/-koi is used to denote action that is durative and intermittent (for a long time, continually, constantly, permanently); frequentative (often, many times, again and again, repeatedly); or intensive (intently, persistently, steadily). Here are some examples: alahai “telling many times;” jonkoi “reminding again and again, constantly;” tehei aliyambi “someone waits sitting for a long time.” Presumably, this form originated from the perfect participle followed by the marker of the genitive case: -hai/ -hei/-hoi < -ha/-he/-ho + i = GEN. Serving as an adverb of manner in a sentence, the perfect participle was regularly followed by the suffix for the genitive. Gradually, adverbial meaning was assigned to this participial form which shifted its category status from participles to converbs. The newly formed converb preserved the perfective meaning by which the perfect participle was characterized but it gained an additional meaning of duration: yasa hada-hai tuwa-mbi eye fix.the.eyes.on-CONV look.at-IMPF “(Someone) stares at (smth./smb.) intently” (ZAKH:194); ere bayan n’alma (ma. niyalma) jui tere muke be daha-me this rich man child that river ACC follow-CONV yabu-hai muke i sekin (ma. sekiyen) de iàina-ha (ma. isina-ha) go-CONV river GEN origin DAT reach-PART “That rich man’s son having gone along the river for a long time, reached the source” (SK:76). The syntactic construction formed by the imperfect and perfect participles in the genitive followed by the word teile “just,” “only,” “alone” is noted by some authors. The meaning of this construction is rendered as “to the extent of ... ”: mutere-i teile “with all one can,” “with all one’s capabilities,” “to extent of one’s power;” jabduha-i teile “as time permits,” “as opportunity allows” (Zakharov, 1879:194;
gorel-2.p65
281
4/17/02, 10:59 AM
282
part four
NL:152, 275). There are also constructions formed by repeatition of the same stem, where the first component is represented by the perfect participle and the second by the converb: baha bahai “unintentionally” (baha- “to get,” “to obtain”); nambuha nambuhai “at random,” “by chance,” “unintentionally” (nambu- “to catch,” “to have fall into one’s hand”) (Zakharov, 1879:194; NL:22, 209). To substitute meanings that are rendered by this converb, the words emdubei “continually,” “persistently,” “steadily,” “frequently,” “earnestly” and daruhai “often,” “continually,” “regularly” are often used. 5.7.6. Terminal Converbs The terminal converb in -tala/-tele/-tolo indicates such subordinate action which is going on simultaneously with the principal action. The latter continues until the final completion of the dependent action. Therefore, completion of the subordinate action is that terminal limit up to which the principal action is leading: ala-tala “until telling;” gene-tele “until going;” mana-tala “until full completion,” “until an end” (mana- “to come to an end,” “to diminish”); sakda-tala “till old age, until old.” Here are some examples: jing erin o-hobi, geli sin-i boo-de now time become-PAST still you(si/sin- SG)-GEN house-DAT isina-tala baita be tookabu-rakå-n arrive-CONV matter ACC delay-PART(NEG)-INT “Now the time has come, while arriving at your house should we still delay the matter?;” håsun moho-tolo strength be.depleted-CONV
“until depleting one’s strength” (ZAKH:197);
utala aniya o-tolo, heni majige ba-ngkakå
so.many.as.this year be-CONV at.all a.little be.tired-PART(NEG)
“So many years have passed, but he has not tired at all” (ZAKH:197);
muse eigen sargan emu jalan-de we(INCL) husband wife one world-DAT dosu-mbu-me (ma. dosimbume) sagda-tala banji-mbi enter-CAUS-CONV get.old-CONV live-IMPF “Until (we) grow old we, the husband and the wife, shall live in the same world;”
gorel-2.p65
282
4/17/02, 10:59 AM
283
morphology
emu bia (ma. biya) o-tolo bou (ma. boo) i dorgi de one month pass-CONV house GEN inside DAT emhun te-fi alone sit-CONV “Until one month had passed [he] sat at home alone” (SK:76). The terminal converb is not used very often in classical Manchu and can be considered an old form. Its use is mainly limited to the syntactic words isitala “up to,” “until” and otolo “till,” “until,” which were originally converbs formed from verbs isi- “to reach,” “to come to,” “to arrive” and o- “to be,” “to become.” This formally analogous converb is preserved in all TungusManchu languages. Presumably, the suffix -tala/-tele/-tolo consists of two components: -ta/-te/-to and -la/-le/-lo. The second component can be considered as the suffix of the locative case -la/-le/-lo. In Manchu this suffix remains only in the structure of converbal suffixes -tala/-tele/-tolo and -nggala/-nggele/-nggolo. In Manchu there is another converb which includes the element -ta/-te/-to, i.e. converb formed by the suffix -tai/-tei/-toi which is also combined from two elements: -ta/-te/-to and i. Attached to a verbal stem, this suffix denotes an extreme degree of an action: ergele-tei gama-ha force-CONV take-PART “(One) took by force;” buce-tei afa-mbi die-CONV fight-IMPF “(One) fights until death;” ejen be weile-re-de beye-be waliya-tai emperor ACC serve-PART-DAT self-ACC throw-CONV o-me mute-re amban kai be(AUX)-CONV be.able-PART high.official COP “When serving the emperor (one) is an official who is able to sacrifice (throw) oneself” (ZAKH:194). 5.7.7. Converbal Forms in -nggala/-nggele/-nggolo The converb in -nggala/-nggele/-nggolo denotes the subordinate action before which the main action takes place. Here are some examples:
gorel-2.p65
283
4/17/02, 10:59 AM
284
part four
dosi-nggala asuki isibu-mbi enter-CONV noise make-IMPF “Before entering (one) makes a noise;” gisun waji-nggala yasa-i muke far seme eye-he speech finish-CONV eye-GEN water in.large.quantity flow-PART far seme “in large quantity” (NL:83); “Before (one) finished speaking, one’s eyes streamed with tears in large quantity” (ZAKH: 198). This converb, like the others, may be derived from certain verbs in their auxiliary function, chiefly from the verb o- “to be,” “to become:” onggolo “before.” All such forms show a tendency to shift to another grammatical class of words, viz. into conjunctions: isitala “up to,” “until;” onggolo “before;” otolo “until;” oci, seci “if;” ofi “because.” 5.7.8. Negative Forms of Converbs Manchu converbs derive their negative forms by specific constructions combined from two components. One of them is represented by a participle in its negative form and another by a corresponding converb derived from the verbs bi- “to be,” “to exist,” o- “to be,” “to become” or se- “to say” in their auxiliary function. The following sentence contains the negative form of the conditional converb: te gaija-rakå o-ci, gåwa gai-ka now take-PART(NEG) be(AUX)-CONV other take-PART de, sin-i dolo ume ehe gåni-he DAT you(SG)-GEN=your inside NEG amiss think-PART ehe guni- “to take something amiss” (NL:71); “If you don’t take (it), and someone else takes, then don’t take this amiss” (PASH2:126). The negative form of the conditional converb is represented here by the construction gaijarakå oci “(literally: “if one doesn’t take”). In the following sentences the negative form of the conditional converb is formed by the negative form of the participle derived from an autonomous verb and the conditional converb derived from the auxiliary verb o-: gene-rakå o-ci, uthai gene-rakå seme hendu go-PART(NEG) AUX-CONV then go-PART(NEG) that say(IMP)
gorel-2.p65
284
4/17/02, 10:59 AM
285
morphology
“If (you are not intending to) go, then say that (you) will not go” (ZAKH:199; QW); ài (ma. si) agda-rakå o-ci tulergi de you believe-PART(NEG) be(AUX)-CONV outside DAT bou (ma. boo) i torgome (ma. torhome) yabu-me tua (ma. tuwa) house GEN around go-CONV look-IMP “If you do not believe (me), go and look outside, around the house;” ere buce-rakå o-ci min-i this die-PART(NEG) become-CONV I(bi/min-)-GEN=my b’ai (ma. biya-i) elden sourin (ma. soorin) àira-me (ma. sira-me) the.moon-GEN light throne inherit-CONV muta-rakå be.able-PART(NEG)
“If he dies, my son, the Light of the Moon, will not be able to inherit
the throne” (SK).
5.8. Finite Verbal Forms As already mentioned in 5 & 5.1, in Manchu all verbal forms can be divided into finite and non-finite ones. Non-finite verbal forms have already been described in previous sections dedicated to participles and converbs. Finite verbal forms serve exclusively as finite predicates which are predicates of simple sentences or predicates of principal clauses of complex sentences. In the Manchu language all finite forms can be ascribed to one of the following moods: the indicative, the imperative, the optative, and the prohibitive. The grammatical category of mood denotes the attitude of the action, rendered by a verb, to the reality from the point of view of the speaker. The category of mood is a morphological (grammatical) means of expression of modality. The grammatical meanings of forms of the mood arose from their use in speech which is the earliest and most important functional form of language and which implies the presence of a speaker, a person who is speaking or writing and who is the central element (or factor) responsible for organizing cognitive and communicative content. This person not only indicates the action as a fact of reality (the indicative, fact-mood)) but evaluates this action as desirable (optative mood), or possible (potential mood), or supposed (suppositional subjunctive mood), etc. This evaluation renders the subjective attitude of a person who is
gorel-2.p65
285
4/17/02, 10:59 AM
286
part four
speaking to the action as a fact of the reality. Each language has its own set of moods. Since each language should indicate an action as a fact of reality all languages possess the category of the indicative which is opposed to the other moods (oblique moods). In other words, the opposition between the indicative and oblique moods is the semantic basis of the grammatical category of mood. 5.8.1. The Indicative. Finite Forms Ascribed to the Indicative Written Manchu monuments reflect a stage of the language when the indicative forms had not been finally formed. Genesis of most forms ascribed to the indicative seems to be transparent. All indicative forms arose from non-finite verbal forms, participles and converbs. Inconsistancy in the way these forms are written—sometimes their components are written separately, sometimes they are written as one—confirms the fact that they are fixed in the process of their development. Most specialists assign the following forms to the indicative: 1) The form in -mbi.
This suffix is attached directly to a verbal stem. Zakharov believed
that the suffix -mbi is a result of the combination of two elements,
i.e. the suffix of the imperfect converb -me and the copula bi: -mbi < -me + bi (Zakharov, 1879:173). It should be mentioned that there are different opinions about the meaning of this form, and no one is universally accepted. According to some manchurologists, the suffix -mbi is used to express the present tense (Zakharov, 1879:173; Mo�llendorff, 1892:9). The others consider that the form in -mbi is used to indicate both the present and future therefore it can rightfully be called the form of the present-future tense (Lebedeva & Gorelova, 1994:78). There is an opinion according to which the suffix -mbi is defined as a marker for the aorist, an aspectual-temporal category (Haenisch, 1986:51). D. Sinor considers that the -mbi marker is indefinite. In his opinion, -mbi does not represent the action in time and does not indicate whether the action is accomplished or not (Sinor, 1968:268-9). According to the most widespread point of view, the form in -mbi is considered to be an imperfect finite form (Norman, 1974:168-9). I will refer to this form as the imperfect finite form. The imperfect form has the following grammatical meanings:
gorel-2.p65
286
4/17/02, 10:59 AM
287
morphology
a) it is used to indicate usual and frequentative or abstract actions. This meaning is close to that of the generic time: indahån dobori tuwahiya-mbi, coko erde håla-mbi dog night guard-IMPF chiken early.in.the.morning sing-IMPF “A dog keeps guard at night, a rooster crows early in the morning” (ORL:193); muke de do-mbi nimaha je-mbi water DAT alight-IMPF fish eat-IMPF “(A bird?) alights on the water, eats fish” (ORL:195); inenggi elben gana-mbi, dobori futa muri-mbi day thatch take-IMPF night rope weave-IMPF “In the day-time (people) gather thatch, at night (people) weave ropes (usually)” (PASH2:31); b) it is used to denote a concrete action which is synchronous to the moment of speech: saman hehe ainu yobodo-mbi? shaman woman why joke-IMPF
“Shamaness, why are you kidding around?” (NSB:59);
si aibi-de gene-mbi?
you(SG) where-DAT go-IMPF
“Where are you going to?” (PASH2:2);
ai baita yabu-mbi? what matter do-IMPF “What are (you) doing?” bi tere buhu-i (ma. buha) soko (ma. suku) weihe be gaji-me I that bull-GEN pelt horn ACC bring-CONV gene-mbi go-IMPF “I am going to bring the pelt and horns of that bull” (SK:78); c) the form in -mbi is used to indicate the future tense: jai uttu o-ho-de, bi usha-mbi-kai further like.this be-PART-DAT I be.angry-IMPF-MDL.PTL “If it goes on like this, I shall be angry” (ORL:193); i inenggi ji-mbi-o jide-rakå-n? he day come-IMPF-INT come-PART(NEG)-INT “Will he come to-day or not?”
gorel-2.p65
287
4/17/02, 10:59 AM
288
part four
urunakå ji-mbi certainly come-IMPF “(He) will certainly come” (ZAKH:178); tere han-de bi beye aca-me gene-mbi that khan-DAT I myself meet-CONV go-IMPF “I myself shall go to meet that khan” (SK:78); d) there are some instances when the form in -mbi indicates such action which has been completed prior to the moment of speech and consequences related to this are relevant to the present moment: s�un dekde-re ergi uce be tuci-mbi sun rise-PART side door ACC appear-IMPF “A rising sun has appeared in the doorway” (NSB:70-1). The form in -mbi normally serves as a predicate of a simple sentence or a predicate of a principal clause of a complex sentence: bi hergen be ara-mbi I character ACC write-IMPF “I am writing characters” (ZAKH:159); nenden amagan be sa-ha-de, doro de hanci o-mbi the.past future ACC know-PART-DAT Tao DAT near becomeIMPF “When (you) know the past and the future, (you) will approach to the Tao” (PASH2:30). Sometimes the form in -mbi occurs in a subordinate clause when dependence between two events (situations) of the outside world is expressed analytically, by converbal or participial forms of the verb se- “to say” in its auxiliary finction (seme, seci “in order”): bi goro ba-ci ulin bota-mbi (ma. buta-) I distant place-ABL property catch-IMPF seme ji-he say(AUX)-CONV come-PART “I came from a distant place in order to gain property (make a fortune);” ama jui we ya gene-mbi se-ci ishunde father child who who go-IMPF say(AUX)-CONV each.other joboco-mbi worry-IMPF
gorel-2.p65
288
4/17/02, 10:59 AM
289
morphology
“The father and the son are worrying about each other (being unable) to deside which of them will go” (SK:78). Complex predicates identical to the imperfect form from the semantic and functional points of view, could serve as the proof of the hypothesis that the form in -mbi originated from the imperfect converb in -me and the copula bi-, for example: juwe niyalma sefu i emgi nure omica-me two man teacher GEN together wine drink.together-CONV bi COP “Two men together with the teacher drink wine” (PASH2:274; LD); yarg’an’i (ma. yargiyan-i) gaika-me really praise-CONV fergue-ma (ma. ferguwe-me) bi be.astonished-CONV COP “(He) is really very much astonished” (SK:78). In such cases the copula bi obligatorily occurs as a separate word since only one syntactic word may follow homogeneous parts of a sentence. The copula bi is always placed after the last autonomous word, however, it is not fused with this verbal form because of actual relations with another verbal form. The copula bi may follow all verbal, participial, and converbal forms. It shows a tendency to combine with autonomous words. Norman mentions two imperfect finite forms in -mi and -m� in the dialect he described. He believes that both imperfect forms correspond to the Literary Manchu form in -mbi and that the form in -mi is functionally identical to the literary form. Nevertheless, there is a difference between the two forms though its character is not quite clear. In Norman’s opinion the elder members of the Sibe family, which provided his language material, used the form in -mi more frequently than the younger members of the same family, especially in narration (Norman, 1974:169). 2) The form of the perfect participle in -ha/-he/-ho, -ka/-ke/ -ko, -ngka/-ngke/-ngko is often used to indicate an action which took place in the past. Here are some examples: deo bohori elan i ba-de younger.brother proper name name of place GEN place-DAT
gorel-2.p65
289
4/17/02, 10:59 AM
290
part four
te-he live-PART “(His) younger brother Bohori lived in Elan place” (GAB:91); juwe di be jafa-ha two god ACC take.prisoner/capture-PART “(He) captured two gods (emperors?)” (ORL:176); hehe heni tani jobodo-ho woman only joke-PART
heni tani “a bit,” “a little,” “only” (NL:128);
“The woman was only joking” (NSB:60).
The majority of specialists ascribe this form to the indicative,
however, I do not support this view due to the fact that the form is
a participle in one of its main functions— that of a predicate of a
simple sentence (see 5.6.1).
3) The form in -habi/-hebi/-hobi and -ngkabi/-ngkebi/-ngkobi.
This form is used to indicate the past indefinite. As for the origin of
this form, it seems to originate from two components, the suffix of
the perfect participle -ha/-he/-ho, -ngka/-ngke/-ngko and the copula
bi: -habi/-hebi/-hobi < -ha/-he/-ho + bi. Here are some examples:
malta de oforo i sangga juwe yasa-i sidende hippopotamus DAT nose GEN hole two eye-GEN in.between banji-habi become-PAST
“The nostrils of the hippopotamus were situated between its eyes”
(ORL:195)10;
nimanggi i elden de bithe håla-habi snow GEN light DAT book read-PAST “(One) read books by the light (reflected) by snow” (ORL:196); absi sin-i coko indahån be min-i håla-ra why you(SG)-GEN chicken dog ACC I(bi/min-)-GEN call-PART sasa amasi foro-me gene-hebi together backward turn-CONV go-PAST “Why did your chicken and your dog turn back as soon as I called them?” (NSB:59); 10 The word malta is translated by Norman as “beluga,” “white whale,” “dol phin” (NL:192). Zakharov considered this word as a name for hippopotamus and also for a mythical creature which looks like a hippopotamus (ZAKHL:872).
gorel-2.p65
290
4/17/02, 10:59 AM
291
morphology
juwe dalba-de juwan funce-re haha-si ili-habi two side-DAT ten be.in.excess-PART woman-PL stand-PAST “More than ten women stood on both sides” (NSB:134); tere bayan n’alma-i (ma. niyalma i) mou (ma. moo) olho-me that rich man-GEN tree dry.up-CONV buce-hebi die-PAST “Having dried up, that rich man’s tree died;” tere ba-de emu amba gurun’i ejen bi-hebi that place-DAT one big state-GEN khan be-PART “In that place the ruler of a big state lived” (SK:79). This form has its negative counterpart: alahabi—alahakåbi (ala- “to tell’, “to report”); jihebi—jihakåbi (ji- “to come”); jonohobi—jonohakåbi (jono-“to mention”). Since the copula bi without any additional suffixes may be used only as a component of a finite form, its fusing with the suffix of the perfect participle (and the imperfect converb) brought into being the verbal forms which serve as predicates of a simple sentence or a principal clause of a complex sentence. In Literary Manchu the copula bi and the participial suffix -ha/ -he/-ho are often written as one, although these are exceptions to this. When the copula bi follows the perfect participle as an independent lexical unit we are dealing with the analytical form of the past tense. In this case one can notice the living relationship between the developing verbal form of the past tense in -habi/-hebi/-hobi and the older analytical form expressed by the construction “the perfect participle + the copula bi.” The copula bi may be replaced by the copula kai which is sometimes written separately from the verb. Thus, instead of the form alahabi the combination alaha kai occurs. As finite forms, both the imperfect in -mbi and perfect in -habi/ -hebi/-hobi, are used in the predicate position. When the autonomous component for one of them is expressed by the converb and for the other by the participle, we have every reason to claim that non-finite forms are the oldest in Manchu and that the forms of the indicative mood have developed from converbs and participles. Norman states that the perfect finite form has the suffix -hi, and for some verbs -qi in the dialect he has studied. It is obvious that this from originated from the known form in -habi/-hebi/-hobi: -habi > -havi > -hai > -hi (Norman, 1974:169; Gorelova, 1986:313).
gorel-2.p65
291
4/17/02, 10:59 AM
292
part four
4) The analytical form Tv-ha/-he/-ho bihe or -ngka/-ngke/-ngko bihe.
This form is constructed analytically, by the suffix of the perfect
participle and existential verb bi- “to be,” “to exist,” followed by the
same suffix of the perfect participle. This analytical form has the
meaning of the past perfect: alaha bihe “(one) had told;” araha bihe
“(one) had written;” genehe bihe “(one) had gone;” gisurehe bihe “(one)
had said;” jongko bihe (one) had mentioned;” tehe bihe “(one) had sat,”
etc. The second element of this form, bihe, may be followed by the
copula bi or kai: alaha bihebi, alaha bihe kai. These constructions occur
at the end of speech (or syntactic period).
5) The form in -mbihe.
This form is used to indicate actions which took place in the past.
Some specialists specify this tense as the past indefinite: alambihe “(one)
told,” genembihe “(one) went;” obombihe “(one) washed” (Zakharov,
1879:174; Pashkov, 1963:36). Others consider this tense to be the
past continuous: arambihe “(one) was writing” (Mollendorff, 1892:10).
Zakharov suggests that this form originated from the imperfect converb in -me derived from an autonomous verb and the perfect participle derived from the verb bi-: -mbihe < -me + bihe. He lists several verb forms: alambihe < alame bihe; genembihe < geneme bihe; obombihe < obome bihe (Zakharov, 1879:174). Here are some examples: malaha boo-de weile-mbihe hat house-DAT make-PAST “(They) made hats (usually) at home;” bi kemuni ere-be niyalma de ere-mbihe I still this-ACC people DAT rely-PAST “I still relied on this man” (PASH1). 6) The form in -mbihebi. As for the origin of this form, it seems to originate from the known form in -mbihe and the copula bi which are written as one in most cases. According to Pashkov, the form in -mbihebi is used to indicate usual and frequentative actions which took place in the distant past (Pashkov, 1963:36). In some cases the copula bi is written separately from the preceding verb in -mbihe. Here are some examples: håwangheo beye nimala fata-mbihebi empress themselves mulberry.tree pick-PAST “It used to be that empresses themselves picked berries from the mulberry tree;”
gorel-2.p65
292
4/17/02, 10:59 AM
293
morphology
bithe be damjan i ujan de lakiya-fi håla-me book ACC carrying.pole GEN end DAT hang-CONV read-CONV yabu-mbihe bi walk-PAST COP “(He) used to walk reading a book which he hung up to a carrying pole” (ORL:194). The analytical form TV-me bihebi probably has the same meaning: nadan tanggå anggala emu ba-de je-me bi-hebi seven one.hundred persons one place-DAT eat-CONV be(AUX)PAST “Seven hundred persons used to eat in the same place (at the same table)” (ORL:95). 7) The form of the imperfect participle in -ra/-re/-ro is used to denote the future: gebu algin duin dere de hafuna-ra fame fame four direction DAT form.free.passage.to.another.placePART “Fame will be spread in four directions (the four cardinal points);” bi sin-de emu sain arga tacibu-re I you(si/sin- SG)-DAT one good method teach-PART “I will teach you a good method” (ORL:180); be damu ging hecen de isina-ha manggi we(EXCL) only capital(city) city DAT arrive-PART after bu-re dabala give-PART MDL.PTL
“We shall give (smth.) back only when we arrive at the capital”
(PASH1:93).
The form in -ra/-re/-ro may be used also to indicate the present:
haha niyalma baita bi-ci teng tang seme jabu-re
man man business be-CONV firm do-PART
tang seme “hard,” “firm,” “fluent;” teng seme “hard,” “firm,” “solid;”
teng tang seme “equally matched,” “straightforward” (NL:271,276);
“If men have business (to do) they do it firmly” (PASH1:75).
8) The form of the optative in -ki is also used to indicate the future.
It is obvious that this future is loaded with the modal meaning of
desire or intention (one is going to do smth.), for example:
gorel-2.p65
293
4/17/02, 10:59 AM
294
part four
muse idu banjibu-ki we(INCL) a.turn form/set.up-OPT “We are going to set up a turn;” fa håbala-ki window paste.paper.over.a.window-OPT “(I) am going to paste paper over windows” (ORL:173); urunakå ujen baili isibu-ki necessarily serious kindness repay-OPT “(I) will repay a serious kindness necessarily” (PASH1:58). 9) There is an analytical form which is used to indicate the future. This form consists of the imperfect participle and the form of the optative derived from one of the functional verbs bi-, o- or se- in their auxiliary functions: alara biki “(one) will tell;” genere oki “(one) will go;” dosoro seki “(one) will bear.” It seems that this form is also characterized by the meaning of an intention or desire. According to Zakharov, this form has a strongly pronounced stylistic colour. It is used mostly in dialogue being replaced in written speech by the form in -mbi in combination with the adverbs urunakå “certainly,” “surely,” “for sure,” “necessarily,” and toktofi “certainly,” “surely,” “without fail” (Zakha rov, 1879:178; NL:299). Here is an example: si amagan inenggi urunakå håwas�a-mbi toktofi you(SG) later day surely develop-IMPF certainly simne-me dosi-mbi examine-CONV succeed.in.an.examination-IMPF amagan inenggi “later,” “a later day;” “Later you will surely perfect yourself and certainly succeed in an examination” (ZAKH:178). Having observed all these forms, one can come to the conclusion that most of them cannot be described as the indicative. Only three forms can be specified as fact-mood, viz., the imperfect finite form in -mbi, the perfect finite form in -habi/-hebi/-hobi, -ngkabi/-ngkebi/ -ngkobi, and the past indefinite finite form in -mbihe. Pashkov suggests ascribing to the indicative one more form in -mbihebi which is used to indicate frequentative actions in the distant past (Pashkov, 1963:36). The origin of these forms is quite clear, and all of them have their own analytical counterparts: -mbi < -me + bi; -habi/-hebi/
gorel-2.p65
294
4/17/02, 10:59 AM
295
morphology
-hobi < -ha/-he/-ho + bi; -mbihe < -me + bihe; -mbihebi < mbihe + bi. Nevertheless the meanings of these forms require further investigation. 5.8.2. Oblique Moods. Optative. Imperative. Prohibitive The indicative is opposed to the oblique moods which designate different attitudes of the person speaking to what is being said (statement, dictum). The diversity of paradigms of the oblique moods stems from the variety of these relationships. Oblique moods are used to denote deontic modality. According to grammatical tradition, the following oblique moods can be establish for Manchu: the optative, the imperative, and the prohibitive. 5.8.2.1. Optative The suffix for the synthetic form of the optative is -ki. This form indicates an action which the person speaking has a desire or an intention to perform or to be performed: bi geneki “I will go” (with the shade of affirmation), si geneki “you would go” (not in the meaning of a strong order but in the meaning of a desire), i geneki “let him/ she go.” Since the optative denotes actions which are not realized, which are to take place in the future, it is inevitably related to the expression of the future. However, the relationship of the optative with the future is a secondary feature which results from the modal nature of this form. That is why the optative forms are often used to indicate the future, and the form in -ki is considered by most specialists as one which denotes the future indicative. When the optative is referred to the second and the third persons (singular and plural) its meaning is very close to that of the imperative. There is a structural reason against analyzing the optative forms for the second and the third persons as the imperative. These forms would become separated from the form for the first person, and the paradigm of the optative would disintegrate. The imperative meaning of the optative occurs only in specific grammatical conditions when reference is made to the second and the third persons. In such cases the imperative meaning arises as additional and secondary to the meaning of the optative. The optative does not possess the category of tense. Here are some examples on the use of the optative: muse aca-hakå-ngge kejine we(INCL) get.together-PART(NEG)-NR a.long.time
gorel-2.p65
295
4/17/02, 10:59 AM
296
part four
goida-ha, bi dosi-fi majige te-ki last.for.a.long.time-PART I enter-CONV a.little sit.down-OPT “We have not seen each other fore a long time; I will come in and sit down for a while” (PASH2:191: MOLL:47); daritai se min-i sargan jui be older.relative PL I(bi/min-)-GEN=my female child ACC inu majige tuwa-ki also a.little see-OPT “Let older relatives also see my daughter” (ORL:172); muse te bou (ma. boo) i baru muda-ki we(INCL) now house GEN towards return-OPT “Let us return home now;” ài (ma. si) ere ba-de banji-me cihala-ci you(SG) this place-DAT live-CONV wish-CONV men’i ere abka i ba-de we(be/men- EXCL)-GEN this heaven GEN place-DAT sasa banji-ki together live-OPT “If you wish to live in this place, live together with us in heaven;” tere ajige faha be na de caci-ki that small seed ACC earth DAT throw-OPT “Throw these small seeds on the earth” (SK:80). 5.8.2.2. Imperative The imperative denotes a direct desire (request or order) of the speaker to cause the interlocutor to perform an action. Since the imperative designates actions which have not begun, it is entirely contained in the grammatical space of the future, however, the imperative does not possess the grammatical category of tense. Participants of an imperative speech act may include besides the speaker and the listener (or listeners) a third person (or persons) from whom a reaction is expected. Consequently, in Manchu there are a number of forms which denote a speaker’s will (desire, order) addressed to both the second and the third persons. In Manchu, the imperative forms are closely related to the expression of honorific meanings. Specialists usually distinguish five forms differing with respect to the honorific meaning they possess. 1) The most commonly used form for the imperative coincides with the verbal stem: ala “tell,” gene “go,” se “say,” te “sit down,” wa “kill.”
gorel-2.p65
296
4/17/02, 10:59 AM
297
morphology
It denotes an address to the second person (singular or plural) occupying a lower or similar position in respect to the person who is speaking. In comparison with the other languages of the Manchu-Tungus linguistic community, only Manchu (and its Sibe dialect) has the imperative form which is rendered by the simplest verbal form— the verbal stem. However, one should keep in mind that this imperative form may include voice and aspectual suffixes: tebu “be seated,” “to settle;” tuwas� ata “look after smb.” Here are some examples: tacikå-i ju-se, min-i gisun be donji
school-GEN child-PL I(bi/min-)-GEN word ACC listen(IMP)
tacikåi juse “pupils,” “students;”
“Pupils! Listen to my words!” (“Listen to me”) (PASH1);
si gai-ki se-ci, uthai gaisu
you(SG) take-OPT say(AUX)-CONV then take(IMP)
Tv-ki se-(aux)- “to want to do smth.;”
“If you want to take (smth.) then take (it)” (PASH2:126);
ama eme be saikan’i gingule-me (ma. ginggule-me)
father mother ACC properly respect-CONV
banji live(IMP)
“Live properly, respecting (your) mother and father;”
mim-be geli sasa makta
I(bi/min-)-ACC too together throw(IMP)
“Throw me together (with smb.) too” (SK:81).
Used as the imperative, non-suffixal verbal stems may denote the
following meanings:
a) an order:
si majige ili you(SG) a.little stand(IMP) “You stand still for a while;” umiyesun be golbon de lakiya belt ACC clothes.rack DAT hang.up(IMP) “Hang up the belt on a clothes rack;” etuku be saikan bukda clothing ACC properly fold(IMP) “Fold the clothing properly” (ORL:166);
gorel-2.p65
297
4/17/02, 10:59 AM
298
part four
suwe hådun baihana-me gene you(PL) quickly go.to.search-CONV go(IMP) “You go and search for (her) at once!” (NSB:22). b) an order with an overtone of advice: dobori deri hålha holo be serems�e night every thief fraudster ACC beware(IMP) “Beware of thieves every night;” aniya-dari bisan hiya be serems�e year-every flood drought ACC beware(IMP)
“Beware of flood and drought every year” (ORL:166);
c) a request or commission:
min-i funde emu jasigan ara
I(bi/min-)-GEN for one letter write(IMP)
“Write me a letter;”
d) necessity or obligation:
min-i gånin de suwe-ni dolo ama-ta
I(bi/min-)-GEN=my opinion DAT you(PL)-GEN inside father-PL
o-ho niyalma ju-se be targabu,
become-PART man child(jui)-PL ACC admonish(IMP)
ahå-ta o-ho niyalma
elder.brother(ahån)-PL become-PART man
deo-te targabu
younger.brother-PL admonish(IMP)
“In my opinion, those (of you) who have become fathers, should
admonish (their) children, and those who have become elder brothers,
should admonish (their) younger brothers” (ORL:167);
e) the imperative can be used to refer to actual events, states, etc.
which are unexpected for the speaker. In other words, it can be used
as an indicative form loaded with the meaning of suddenness:
gaitai kiyatar seme emgeri akjan akja suddenly rumbling(ONOM) once a.thunder thunder(IMP) “Suddenly thunder burst out (with the sound of rumbling)” (ORL:165). One can find sentences where a non-suffixal form denotes strong intention on the part of the speaker to perform an action. Thus, it
gorel-2.p65
298
4/17/02, 10:59 AM
299
morphology
is possible to classify this usage of the imperative as the categorical future: bi boo-de cai fuifu-fi sim-be aliya I house-DAT tea boil you(si/sin- SG)-ACC wait(IMP) “Having boiled tea I shall wait for you at home”(PASH1). A number of verbs have an irregular form of the imperative. Some verbs are formed by the suffixes so and su: oso (o- “to be,” “to become”); baisu (bai- “to ask for,” “to seek for,” “to wish”); bisu (be“to be”); gaisu (gai- “to take”); takasu (taka- “to know a person,” “to recognize,” “to identify”). It is interesting to mention that the suffix -su is used to indicate the imperative forms in the Nanai (Gold) language: oso “be,” “become;” gasu “buy”(this verb correlates with the Manchu verb gai- “to take”); disu “come” ( this verb correlates with the Manchu verb ji- “to come”). These Nanai forms are considered to be the most ancient, however, in the contemporary language the new forms are normally used side by side with old forms: odu, gadu, didu (Avrorin, 1961, 124-5). Some imperative forms are formed by the suffixes -fu, -nu or -n: je/jen (je- “to bear,” “to tolerate”); jefu (je- “to eat”); jon (jo- “to mention”); yon (yo- “to do,” “to walk,” “to leave”). Apparently, the verbal form jefu only seems irregular while in reality it is formed according to the general rule. In the imperative, in contrast with the other forms (jembi, jefi, jeke, etc.), the verb jefu has preserved the old stem which is similar to that existing in the Nanai (den- ) and Evenki (jep-/jev-/jevu-) languages (Avrorin, 1961:124; Lebedeva & Gorelova, 1994:81). The imperative form of the verb ji- “to come” is formed by the suffix -u: jiu (presumably, originated from jisu > jiu). The verbs ending in the suffix -nji or -ji terminate in -u in the imperative: alanji- “to come to report”—alanju; benji- “to send (hither),” “to deliver (hither)”—benju; gaji- “to bring,” “to bring along”—gaju. 2) The optative form in -ki is often used as imperative when a person who is speaking treats the person addressed as an equal. This form occurs to soften the imperative attitude: ala-ki tell-OPT “Let (me) speak;”
gorel-2.p65
299
4/17/02, 10:59 AM
300
part four
te-ki sit-OPT “Please, sit down” (ZAKH:179); leole-ki discuss/talk.over-OPT “Let’s talk” (ORL:172). 3) The form in -kini is used to address a person who is lower in social position. It denotes an order to perform an action immediately: alakini “let him speak” or “allow him to speak;” gaikini “let him take” or “allow him to take;” genekini “let him go” or “allow him to go.” Here are some examples: jalan as�s�a-me jakån jui sabu-kini generation move-CONV eight child see-IMP “Moving generations, let him see eight children;” sergådai ere-ci amasi ninju aniya nimeku akå proper name this-ABL after sixty year illness there.is.not tanggå aniya targa akå ujun ju-se uji-kini hundred year abstain there.is.not nine child(jui)-PL raise-IMP “Sergådai, after this, will not be ill for sixty years and will not abstain for a hundred years. Let him raise nine children” (NSB:58). According to Zakharov, the form in -kini consists of the suffix of the optative -ki and the interrogative particle ni. It is used in respect to the second or the third persons in order to cause the addressee to act or fulfil a command immediately: yabu-bu-kini carry.out-PASS-IMP “(They) are ordered to carry out;” sin-i ciha i o-kini you(si/sin- SG)-GEN desire GEN be-IMP “Let it be in accordance with your desire” (ZAKH:180); suwe gene-kini, mim-be ume dabu-re you(PL) go-IMP I(bi/min-)ACC NEG press.to.come-PART “You should go, but not press me to come” (ORL:172). According to Lebedeva, the form in -kini is used when the person addressed is to cause another person (persons) to perform an action: okini “order him (or them) to become,” gajikini “order him (them) to
gorel-2.p65
300
4/17/02, 10:59 AM
301
morphology
bring.” In her opinion, the suffix -kini consists of two components. The first is represented by the marker of the optative in -ki, and the second component ni indicates the third person. It originated, as in other languages of the Tungus-Manchu family, from the pronoun of the third person in the genetive: i “he” + (n)i > ini. Here are some example: hehe n’alma (ma. niyalma) oci gemu deijiku gaji-me woman person TOP all firewood bring-CONV ji-kini come-IMP “As far as women are concerned, let them come bringing firewood;” ere daifu i sargan mergen erdemu be dahame this doctor GEN wife wise powerful ACC since men’i jue (ma. juwe) n’alma (ma. niyalma) we(be/men-EXCL)-GEN=our two man de sargan o-kini DAT wife be-IMP “Since this doctor’s wife is very wise, let her be a wife to both of us” (SK:81). According to Norman, the form in -kini is used as the optative in the Sibe dialect. It is used to express a desire or wish concerning the third person (Norman, 1974:172). 4) The form in -cina/-kina is considered to be an imperative form by most specialists (Zakharov, 1879:180-81; Pashkov, 1963:37). The form in -cina/-kina is used by the speaker when he addresses someone who is lower in social position. The accomplishment of the action is considered by the speaker as desirable: alacina (alakina) “speak out (if you wish);” tuwacina (tuwakina) “have a look (if you wish),” etc. According to Zakharov, this form originated from the combination of the conditional converb in -ci and the interrogative particle na: -ci + na > -cina. In the author’s opinion, this form expresses a polite request formulated as a question. The meaning rendered by the form in -cina can be translated as the following words and phrases: “please,” “very likely,” “let it be,” “let it be in this way,” “isn’t it so?,” “isn’t it true?”. The word form bicina (the form in -cina derived from the verb bi- “to be,” “to exist”) can be translated as “let it be,” “let it be in this way!,” “isn’t it true?”. The word form secina (the form in
gorel-2.p65
301
4/17/02, 10:59 AM
302
part four
-cina derived from the verb se- “to say”) means literally “let me say.” It is very often used as a note of exclamation at the end of an utterance in the meaning “this way it is!,” “so it goes!” (Zakharov, 1879:180-1). Here are some examples: uba-de te-cina place-DAT live-IMP “Please, live in this place;” kemuni ji-cina often come-IMP “Please, come often” (ORL:174). si ainu hendu-rakå, min-de daljakå you why say-PART(NEG) I(bi/min-)-DAT unconnected.with se-he bi-cina say-PART be-IMP “Why didn’t you explain, but said that I had nothing to do with this;” ajige ju-se damu sefu i jilgan be donji-ha de, small child(jui)-PL only teacher GEN voice ACC hear-PART DAT fayangga gemu tuhe-mbi se-cina soul all fall-IMPF say-IMP “As soon as little children hear the teacher’s voice, all [of them] lose courage [at once], so it goes (literally: let me say)!” (ZAKH:181); damu ai emgeri sin-de bai-ki se-he be only what one you-DAT ask-OPT say(AUX)-PART ACC daha-me, si uthai te-de bu-cina follow-CONV = since you then that(tere)/he-DAT give-IMP “Since he wants to ask you for smth. for the first time, give (it) to him” (PASH2:117); ài ama-i ici emu mudan t’ua-cina (ma. tuwa-cina) you father-GEN toward one time look.at- IMP “Won’t you have a look at your father?” (SK:83). 5) The interrogative form of the imperfect participle in -rao/-reo is also used as the imperative. This form is used to address older people or those of higher social position (or status). Persons may be addressed by the speaker of equal age or position, but in this case this form is used to express extreme politeness or humility on the part of the speaker. Here are some examples:
gorel-2.p65
302
4/17/02, 10:59 AM
303
morphology
te-re-o sit-IMP(PART-INT)
“Please, sit down!” (ZAKH:180);
si unenggi sa-ra-ngge getuken o-ci, uthai
you(SG) really know-PART-NR clear be(AUX)-CONV then
min-de ulhi-bu-me ala-re-o
I(bi/min-)-DAT understand-CAUS-CONV tell-IMP(PART-INT)
“Actually, if you know (smth.) exactly (literally: clearly), then explain
(it to me), please” (PASH2:80);
gilja-me gama-ra-o forgive-CONV forgive-IMP(PART-INT) “Please, excuse me” (NL:106); sakda eme be gåni-fi haila-mbi old mother ACC think-CONV feel.sympathy-IMPF dulembu-re-o pass.through-IMP(PART-INT)
“Thinking of [my] old mother, please, feel sympathy with me and
let me pass” (NSB:63).
The same form normally serves as the interrogative participle:
bai-re-ngge min-i jui i ajigen ergen
ask.for-PART-NR I(bi/min-)-GEN=my child GEN young life
be aitubu-re-o
ACC revive-PART-INT
“This is what (I) ask: won’t (you) revive the young life of my son?”
(NSB:31).
That is why the form in -rao/-reo may be considered as the
interrogative form of the imperfect participle or the imperative form
depending on its function in the syntactic context.
Norman considers the form in -cina (-cinaa) as the hortative. In his opinion, this form is used to express an exhortation or wish concerning the second person in the dialect he described (Norman, 1974:172). Normally, all synthetic forms for the imperative may be replaced by analytical ones which consist of a converbal form derived from an autonomous verb and an imperative form derived from one of the following functional verbs bi-, o- or se-: ala—alame bisu, alame oso; alakini—alame okini; alarao—alame sereo.
gorel-2.p65
303
4/17/02, 10:59 AM
304
part four
5.8.2.3. Prohibitive By its meaning, the prohibitive mood is the negative imperative concerning the second person (singular and plural). Formally it is distinguished from the imperative which is expressed by zero and coincides with the verbal stem. The prohibitive is formed with the suffix of the imperfect participle -ra/-re/-ro and the interrogative particle ume which precedes the participle: ume genere “do not go,” ume songgoro “do not cry.” Here are some examples: age ume ehe gåni-re elder.brother/sir NEG evil think-PART “Sir, do not take something amiss!” (PASH2:86;QW); ume haira-ra NEG feel.sympathy-PART “Do not feel sorry (for smb.)” (NSB:13); ume helede-re NEG loiter-PART “Do not loiter!” (SK:26); bai-re-ngge ume holto-ro wish-PART-NR NEG lie-PART “This is what I wish: do not lie!” (NSB:59); ài (ma. si) taka ume saci-re bi ebu-ki you for.a.short.time NEG hack-PART I get.down-OPT “You do not hack for a short time; I will get down;” dorgide muduri meihe umuài (ma. umesi) labdu, ài ume inside dragon snake very a.lot.of you(SG) NEG gele-re fear-PART
“There are a lot of dragons and snakes inside, do not be afraid”
(SK:82).
The form for the prohibitive is a special form which stays apart from
other verbal forms in Manchu. The composite character of this form,
especially the form of an autonomous verb, is very close to the
negative forms which can be discovered in other languages of the
Tungus-Manchu family. At the same time the positive form of the
imperative concerning the second person in Manchu is sharply
different from those occurring in other Tungus-Manchu languages.
Presumably, the imperative and the prohibitive forms concerning
the second person developed as a result of totally different processes
and in different periods of time.
gorel-2.p65
304
4/17/02, 10:59 AM
305
morphology
5.8.3. Forms of Epistemic Modality In Manchu, forms of epistemic modality include a variety of grammatically opposed forms which denote different attitudes of the speaker towards the action viewed as real or unreal. To speak about a fact, the speaker uses the indicative forms. To speak about something which is unreal, possible, probable, desirable, potential, etc., the speaker can choose different forms of epistemic modality. Forms of real modality, that is indicative, are mostly expressed synthetically, by simple verbal forms. Forms of unreal modality are formed analytically. The meaning of a desire, wish or an intention to perform an action is expressed by the analytical form which consists of the optative form of an autonomous verb and a form of the verb se- “to say” in its auxiliary function: alaki sembi “I (you/he/she) wish to tell; geneki sehebi “I (you/he/she) wished to go.” Here are some examples: sin-i beye gene-ki se-he dabala, you(si/sin- SG)-GEN self go-OPT say(AUX)-PART merely we sim-be gene se-he who you(SG)-ACC go(IMP) say-PART “You yourself wished to go; who forced you to go?” (literally: “who said to you: go!”) (PASH2:101; QW); bi in-i mudan be dinji-ci, sin-de I he(i/in-)-GEN = his sound ACC hear-CONV you(SG)-DAT aika jaka bai-ki se-re gese something thing ask.for-OPT say(AUX)-PART like “I am hearing his voice; it seems he is asking you for something” (PASH2:116; QW); nure omi-ki se-me gåni-ci nure tuci-mbi wine drink-OPT say(AUX)-CONV think-CONV wine appear-IMPF “If (you) wish to drink some wine, wine will appear;” dobi uncehen be tuki-fi (ma. tukiye-) na be tanda-me fox tail ACC raise-CONV earth ACC beat-CONV kaica-ki se-re-de ... shout-OPT say(AUX)-PART-DAT “When a fox, having raised its tail, is going to beat (with it) on the earth ... ” (SK:85). The meaning of potentiality is expressed by the analytical form which consists of the form of the conditional converb in -ci, derived from
gorel-2.p65
305
4/17/02, 10:59 AM
306
part four
an autonomous verb, and a form of the verb o- “to be,” “to become” in its auxiliary function: seci ombi “(it is) possible to say,” gånici ombi “(it is) possible to think.” The following patterns may also express this meaning: Tv-me baha- and Tv-me mute- (baha- “to be able,” mute“can,” “to be able”). Here are some examples: tondo nomhon niyalma se-ci o-mbi upright docile man call-CONV be(AUX)-IMPF “It is possible to call (him) an upright and docile man;” bi nure omi-me bahana-rakå I wine drink-CONV be.able-PART(NEG) “I am not able to drink wine” (PASH1:74); emu gisun be tucibu-me mute-rakå one word ACC pronounce-CONV be.able-PART(NEG) “(One) is not able to pronounce a word” (ZAKH:134); giyam-be (ma. giyan be) bodo-ci ere sargam-be regulations-ACC follow-CONV this woman-ACC (ma. sargan be) bi gai-ci o-mbi I take-CONV be(AUX)-IMPF “According to the regulations, I can take this woman;” ài min-i sargan o-ci o-mbi-o? you I(bi/min-)-GEN wife become-CONV be(AUX)-IMPF-INT “Can you become my wife?” (SK:85). The analytical form, which consists of the imperfect converb in -me
derived from an autonomous verb and a form of the verb tuwa- “to
look at,” “to see” in its auxiliary function, indicates an attempt to
perform an action: bodome tuwambi “I (you/he/she) try to look at
smth.;” gajime tuwambi “I (you/he/she) try to take smth.”. Here are
some examples:
bi ere dobori bodo-me tua-mbi (ma. tuwa-mbi)
I this night think-CONV look(AUX)-IMPF
“I shall try to think tonight;”
geli emu meyen sende-me (ma. cende-) tua-ki (ma. tuwa-ki)
again one time verify-CONV look(AUX)-OPT
“(We) shall try to verify once more” (SK:86).
An action which must (should, ought to) be performed is also
expressed analytically. This analytical form consists of the form of
gorel-2.p65
306
4/17/02, 10:59 AM
307
morphology
the conditional converb derived from an autonomous verb and a form of the verb aca- “to meet,” “to get together,” “to combine” in its auxiliary function. Here are some examples: ere gisun be niyalma tome kice-ci aca-mbi this word ACC people every study-CONV meet(AUX)-IMPF “All people must study these words” (GAB:23); erin-dari gele-re, olho-ro be tebu-ci time-every fear-PART fear-PART ACC keep.in.one’s.heart-CONV aca-mbi meet(AUX)-IMPF “At all times one ought to keep in one’s heart an apprehension and caution” (ZAKH:188); manju bithe håla-ra niyalma oci, urunakå hergen tome Manchu book read-PART person TOP necessarily letter each gemu getukele-me sa-ci aca-mbi all make.clear-CONV know-CONV meet(AUX)-IMPF “Every person, who studies Manchu, necessarily ought to know every letter exactly” (PASH2:69; QW). The analytical form, which consists of the negative form of the imperfect participle in -rahå (-rahå < -ra + akå) and a grammatical form of the verb se- “to say” in its auxiliary function, normally denotes the meaning of apprehension that an action might take place contrary to a desire or expectation. This meaning is very close to that expressed by the following patterns including connecting devices: “(in order) not to do + inf.;” “(so) that ... may not (might) + inf.;” “in order that ... may not (might) + inf.;” “I am afraid that.” The composite morpheme -rahå may be replaced with the participial suffix -ra, but the meaning of the whole analytical form does not change: alarahå sembi, alara sembi “(I) am afraid that (he) might tell.” Sometimes the negative form of the imperfect participle in -rakå in combination with a form of the verb se- are used to produce the same meaning. Here are some examples which express the meaning “anxious expectation:” ama eme damu nime-rahå se-me jobo-mbi father mother only ill-PART(NEG) say(AUX)-CONV worry-IMPF “I worry that my father and mother might get ill;” banuhån heolen o-jorahå lazy negligent become-PART(NEG)
gorel-2.p65
307
4/17/02, 10:59 AM
308
part four
“(I am afraid) that (lest) (he) might (should) become lazy and negligent” (ZAKH:210); gåwa emu baita de gene-ki se-ci, geli another one business DAT go-OPT say(AUX)-CONV again sim-be ji-de-rakå se-mbi you(si/sin- SG)-ACC come-PART(NEG) say(AUX)-IMPF “(I) wanted to go on another business, but (was afraid) that (lest) you might (should) come (to visit me)” (PASH2:92; QW). The form in -rahå may be replaced by the interjection ayoo “not likely,” “nothing of the kind” which takes a place directly before the functional verb: amus�abu-re ayoo se-mbi become.sleepy-PART not.likely say(AUX)-IMPF
“(I am afraid) that (lest) (I) might (should) become sleepy”11;
damu sain niyalma ji-de-rakå o-jo-rahå,
only good people come-PART(NEG) AUX-PART(NEG)
ehe niyalma geli jide-re ayoo se-mbi bad people also come-PART not.likely say(AUX)-IMPF “(I am afraid) that (lest) good people might (should) not come, and there is also a danger that (lest) bad people might (should) come” (ZAKH:210). To stress the meaning of apprehension both forms are used: ala-rahå ayoo se-mbi tell-PART(NEG) not.likely say(AUX)-IMPF
“(I am afraid) very much that (lest) (one) might (should) tell”
(ZAKH:210).
In negative sentences with the functional verb o-, the negative particle
akå is attached to a participial form of an autonomous verb, whereas
the morpheme -hå (< akå) occurs with grammatical forms of the verb
o- in its auxiliary function:
tuwa-rakå o-jo-rahå see-PART(NEG) AUX-PART(NEG)
“(I am afraid) that (lest) (one) might (should) not see” (ZAKH:211).
11
Norman in his Lexicon gives another analysis of the word form amus� abure, i.e. amu “sleep” + s�abura- “to get sleepy” > amus�abu- “to become sleepy” (NL:16, 256). The word ayoo is considered by him as a final particle denoting fear or doubt (NL:21).
gorel-2.p65
308
4/17/02, 10:59 AM
309
morphology
It seems that the combination of the negative form of the imperfect participle (-rakå) with the word ayoo has the same meaning of “anxious expectation:” bi damu age si jide-rakå ayoo I only sir you(SG) come-PART(NEG) not.likely se-re dabala say(AUX)-PART MDL.PTL “I am only (afraid) that you, sir, might (should) not come (to visit me)” (PASH2:74; QW). The modal meaning of probability can be expressed lexically, with the help of the following words: ainci “perhaps,” “probably,” “apparently,” “presumably;” aise “perhaps;” dere “probably,” “likely.” The first word normally stands in the beginning of a sentence, and the two last ones are placed at the end of it. Very often the word dere is written as one with a verb. The word ainci may be omitted, however either aise or dere are obligatory at the end of sentences expressing modality of probability. Here are some examples: ainci baitala-ci o-mbi dere probably use-CONV be(AUX)-IMPF likely “Probably, it is possible to use (it);” ere gese-ngge oci ainci bi-dere this same-NR TOP probably be/exist-likely/probably
“(Things) similar (to this) are likely to exist;”
min-i boo-de ji-mbi dere seme aliya-hai,
I(bi/min-)-GEN house-DAT come-IMPF probably that wait-CONV
si umai ji-hekå
you(SG) NEG come-PART(NEG)
“I had been waiting for a long time (thinking), that you, probably,
would come (to visit me), but you didn’t come;”
gene-he aise go-PART perhaps “Likely, he has gone away” (ZAKH:209). 5.9. Analytical Verbal Forms and Constructions Analytical verbal forms are very widespread in Manchu. They are used to denote various temporal, modal, and aspectual meanings. I shall use the term “analytical form” to refer to language entities which
gorel-2.p65
309
4/17/02, 10:59 AM
310
part four
are combined from participial or converbal forms derived from autosemantic verbs and finite forms—participles or verbs—derived from the verbs bi- “to be,” “to exist,” o- “to be,” “to become” or se“to say” in their auxiliary functions. Analytical verbal forms are used to indicate predicates in simple sentences or in principal clauses of complex sentences. Most synthetic forms which are ascribed by specialists to the indicative mood, are analytical in origin. These include the imperfect and perfect finite forms in -mbi (< -me + bi) and -habi/-hebi/-hobi (< -ha/-he/-ho + bi). The form in -mbihe is also analytical in origin: -mbihe < -me + bihe (see for details 5.8.1). The alternation of synthetic and analytical variants of these forms is striking evidence of an immediate connection between them. One can observe the variants Tv-mbi and Tv-me bi; Tv-habi/-nebi/-hobi and Tv-ha/-he/-ho bi; Tv mbihe and Tv-me bihe even in the same corpus of texts (possibly, with different shades of meenings). Ascribed to the indicative, the analytical forms Tv-ha/-he/-ho bihe and Tv- ngka/-ngke/-ngko bihe have no corresponding synthetic forms. As it was shown in Section 5.8.3, the forms of epistemic modality are also characterized by analytical structure. In the auxiliary function some more verbs may be used: ali“accept,” “to receive;” baha- “to get,” “to obtain,” “to be able;” goida“to last for a long time,” “to endure;” hami- “to approach,” “to be close;” jafa- “to take in the hand,” “to hold,” “to grip;” s�angga- “to finish,” “to come to an end,” “to be accomplished;” waji- “to finish.” A peculiarity of these verbs is that their lexical meanings are weakened within analytical constructions, but do not vanish entirely. Combined with a main (autosemantic) verb, the verb ali- forms the passive construction. In this case the verb ali- follows the main verb which occurs in the form of the imperfect participle followed by the accusative: han ci banji-re be ali-ha khan ABL be.born ACC get-PART “He was born of a khan” (literally: “He has got (his) birth from a khan”) (PASH:53). The verb ali- in the form of the imperfect converb frequently combines with the verb gai- “to take” preceding it, as in: gosi-re be ali-me gai-mbi love-PART ACC get-CONV take-IMPF “He is loved” (literally: “He is getting a love”) (PASH:53).
gorel-2.p65
310
4/17/02, 10:59 AM
311
morphology
Within analytical constructions the verb baha- “to get,” “to obtain,” “to be able” may occupy the first position, and the main (autose mantic) verb follows it taking certain temporal and modal grammatical formants. In these combinations the verb baha- has the form of the perfect or conditional converbs which are governed by the main verb. The latter gains additional meaning which can be rendered by the verbs “to achieve,” “to attain,” “to gain”: baha-fi sa-ha get-CONV know-PART “(One) succeded in learning (it)” or “(One) managed to learn (it);” baha-fi aca-hakå get-CONV meet-PART(NEG) “(One) failed to meet (smb.)” (PASH:52). Performing in the auxiliary function, the verb baha- may occupy the usual place for auxiliary verbs—that is the position after the main verb. In this case the main verb may occur in the form of the perfect converb: jafa-fi baha grasp-CONV get(PART)
“(One) succeeded in grasping (smth.)” (PASH:52).
The main verb may occur in the form of the imperfect converb, and
the whole construction indicates the potential meaning of an action:
nure omi-me bahana-rakå wine drink-CONV be.able-PART(NEG) “I am not able to drink wine” (PASH:52). The verb goida- “to last for a long time,” “to endure” in the forms of the imperfect converb or the perfect participle in combination with the main (autosemantic) verb indicates that an action expressed by this construction took place long ago. Here are some examples: bi ara-ha goida-ha I write-PART endure-PART “I have written long ago;” donji-fi goida-ha hear-CONV endure-PART “I heard long ago” (PASH:53).
gorel-2.p65
311
4/17/02, 10:59 AM
312
part four
The form of the imperfect converb of this verb shows a tendency to shift its category status from converbs to adverbs as well as the negative form of the perfect participle: goidame “long ago,” “long since,” “for a long time;” goidahakå “before long,” “in a short time.” This tendency is seen in the following example: goidame te-mbi “(One) has been sitting for a long time” (PASH:53). The verb hami- “to be close to,” “to approach,” “to almost reach” in the form of the imperfect converb or the perfect participle in the combination with the main verb has the adverbial meaning “about to,” “on the point of,” “almost”: buce-re hami-me die-PART be.close.to-CONV “being close to death” or “approaching to death;” emu aniya hami-ka one year almost.reach-PART “Almost a year (passed)” (PASH:53). The verb jafa- “to take in the hand,” “to grasp,” “to hold,” “to grip” occupies the first position in the analytical construction. Taking the suffix of the perfect converb, this verb add the meaning “to set (to),” “to get down (to)” to the main verb: fe forgon i niyalma be jafa-fi gisure-ci old time GEN people ACC grasp-CONV talk-CONV “If (we) set to talk about people of the old time ... ” (PASH:52). The verbs s� angga- “to finish,” “to terminate successfully,” “to be accomplished” and waji- “to finish,” “to end” in preposition to the main verb indicate that an action, expressed by it, refers to the plusperfect: waji-fi obu-ha manggi s�angga-me finish-CONV make-PART after be.completed-CONV ara-ha write-PART
“(One) wrote after he had finished what he had made (a matter)”
(PASH:52).
As one can see, these combinations display a smaller degree of
cohesion between their elements because the lexical meanings of
gorel-2.p65
312
4/17/02, 10:59 AM
313
morphology
functional verbs do not disappear entirely. The functional verbs in these constructions partly preserve their lexical meanings in contrast with the analytical forms where the verbs bi-, o-, se- perform a solely auxiliary function. Another kind of combination of autosemantic and functional verbs exists. The distinctive feature of these combinations is that they serve as subordinate predicates, i.e. those language units which are used to indicate predicates in subordinate clauses. In these combinations, the functional verbs normally take converbal or participial suffixes, and participles are followed by certain case markers. I prefer to call such combinations analytical constructions, in contrast with analytical forms. As shown in 5.6.1, participles, perfect and imperfect, display aspectual characteristics rather than temporal ones. Converbs have no grammatical category of absolute tense at all (see for details 5.7). They are characterized by relative (retrospective) tense. It means that their temporal characteristics depend on the tense of finite verbal forms which designate principal actions. Manchu participles and converbs developed specific analytical constructions to denote the whole complex of modal-temporal and aspectual meanings. At the same time they preserved the main function of non-finite forms, participles and converbs, that indicate various relationships (including specific temporal ones if they are the particular meanings of participles and converbs) between subordinate and principal forms and, using them, between subordinate and principal clauses. Usually these constructions consist of two components, the first of which is represented by a finite, or participial, or converbal form of an autosemantic verb, whereas the second element is a participle in a certain case form or a converb derived from one of the functional verbs. Normally, these are the Manchu verbs bi- “to be,” “to exist;” o- “to be,” “to become;” and se- “to say.” These analytical construc tions are very numerous. They have a very diverse structure and semantics. Presumably, the class of analytical constructions is open, i.e. the language may develop new patterns of them. This process of elaboration of new patterns stems from the necessity to find better ways of expressing the diversity of relationships between the events of the outside world which are rendered by principal and subordinate clauses. These analytical constructions are the basis for developing various conjunctive devices including conjunctions proper. Under certain conditions, the functional components of analytical constructions tend to loosen their sentence-bound rigidity and exibit
gorel-2.p65
313
4/17/02, 10:59 AM
314
part four
the ability of functioning beyond the given syntactic context (e.g., in combinations with finite verbal forms and names), gradually developing into analytical markers of the conjunctive variety. This is a characteristic of most Manchu subordinate conjunctions, developed from participial-case and converbal forms, e.g., ojorode, ohode “when;” bici, oci, seci “if;” bicibe, ocibe, secibe “although,” “though;” seme “that,” “in order,” “so that,” “though;” ofi “because.” In this context it is appropriate to point out that the process of separation of the functional component in the analytical construction is easier if this component contains fewer grammatical markers and, therefore, performs fewer functions in the structure of the dependent clause. Thus, in the Tungusic languages—for example, in Evenki— the process of separation seems more problematic in the case of the so-called personal converbs having personal suffixes in their linear structure rather than with so-called simple converbs having no markers of person. In Manchu, this process is easier than in Evenki, since it is facilitated by the specific features of the Manchu morphological basis and the structure of its verbal forms. The following is an incomplete list of analytical constructions in classical Manchu: 1) Constructions where the functional component is expressed by the imperfect converb derived from the verb bi-. The autosemantic component can be expressed by the forms of the imperfect and perfect participles in their positive and negative forms: a) Tv-ha bime utala taci-ha bi-me, kemuni so.many.as.this learn-PART be(AUX)-CONV still ubaliyambu-me mute-rakå translate-CONV can-PART(NEG)
“Having been learning for so many years (one) still cannot translate”
(ZAKH:191);
b) Tv-rakå bime
tere anggala min-i beye s�untuhuni alban de
that not only I(bi/min-)-GEN self all.day duty DAT
ka-me jabdu-rakå bi-me
fulfil-CONV have.time-PART(NEG) be(AUX)-CONV
tere anggala “moreover,” “all the more;”
gorel-2.p65
314
4/17/02, 10:59 AM
315
morphology
“Moreover, since I myself have no time to fulfil (my) duty during all the day long, … ” (PASH2:93; QW). 2) Constructions where the functional component is expressed by the imperfect converb derived from the verb se-. The autosemantic component can be expressed by the forms of the imperfect and perfect participles, in their positive and negative forms: a) Tv-ha seme wesihun foro-ro gånin akå o-ci, upward turn.toward thought there.is.not be(AUX)-CONV tanta-ha se-me, inu ai tusa beat-PART say(AUX)-CONV also what benefit “If there is not a thought which turns (smb.) toward the top, even if (one) beats (him), what is the profit?” (ZAKH:199). b) Tv-rakå seme sin-i baha-ra giyan o-ci, uthai you(SG)-GEN get-PART right be-CONV then tems�e-rakå se-me inu baha-mbi contend-PART(NEG) say(AUX)-CONV also get-IMPF “If you have the right to get (smth.), even if (you) don’t contend, (you) all the same will get (it)” (PASH2:122; QW). 3) Constructions where the functional component is expressed by the converb in -hai/-hei/-hoi derived from the verb bi-: Tv-me bihei uttu banji-me bi-hei ere hehe dergi de so live-CONV be(AUX)-CONV this woman soul DAT gåni-me ...
think-CONV “Having lived for such a long time, this woman thought in her soul ...;” eigen sargan junafi banji-me bi-hei husband wife two.persons live-CONV be(AUX)-CONV “Husband and wife, two persons, lived for a long time … ” (SK:85). 4) Constructions where the functional component is expressed by the conditional converb derived from the verbs bi- , o- , and se-: Tvme bici; Tv-me oci; Tv-ra bici, Tv-ha bici, Tv-ra oci; Tv-rakå oci; Tv-rakå seci; Tv(IMP) seci.
gorel-2.p65
315
4/17/02, 10:59 AM
316
part four
As can be seen from the examples listed above, patterns are different in respect of the form of autosemantic components which they contain in their structure. These autosemantic components may be expressed by the following forms: a) the imperfect converb in -me: Tv-me bici gucu-se-i dolo gucule-re-de, damu ishinde friend-PL-GEN inside make.friends-PART-DAT only mutually mujilen de tebu-me heart DAT keep-CONV
bi-ci, teni gucu i doro dabala
be(AUX)-CONV then friend GEN rule COP(only)
mujilen de tebu- “to be concerned about, to keep in mind” (NL:203);
“When people are friends, they follow the main principle of friendship
only if they care about each other” (PASH2:87; QW);
b) the imperfect participle in -ra/-re/-ro: Tv-ra/-re/-re oci
yaya baita de damu beye-i cihakå ba-be, inu
every matter DAT only self-GEN unwilling matter-ACC also
niyalma de ume isibu-re o-ci,
people DAT NEG repay-PART be(AUX)-CONV
aya-ci ojo-rakå se-re
great(ayan)-ABL be-PART(NEG) say(AUX)-PART
ba uthai akå dere
SBSTR then COP.NEG (there.is.not) probably
“On every occasions, if we don’t do to others what we don’t wish
to themselves, everithing will be fine (I suspect)” (PASH1:102; QW);
c) the negative form of the imperfect participle in -rakå: Tv-rakå oci
bi esi mute-re-i teile
I certainly be.able-PART-GEN to.the.extent.of
fas�s�a-ci talu de sin-i baita be exert.effort-CONV by.chance DAT you(SG)-GEN matter ACC mutebu-ci, si inu ume urgunje-re bring.about-CONV you certainly NEG rejoice-PART sin-i baita be mute-rakå o-ci, you(SG)-GEN matter ACC be.able-PART(NEG) be(AUX)-CONV si inu ume usha-ra you(SG) certainly NEG be.angry-PART
muterei teile “with all one can,” “with all one’s capabilities;”
talu de “by chance,” “in case ... ” (NL:206, 271);
“I will try to do my best; if I accomplish your task, don’t cheer; if
I don’t, don’t be angry with me” (PASH1:106; QW);
gorel-2.p65
316
4/17/02, 10:59 AM
317
morphology
ere buce-rakå o-ci min’i
this die-PART(NEG be(AUX)-CONV I(bi/min-)-GEN=my
b’ai (ma. biyai) elden sourin s�ira-me mute-rakå
moon light throne inherite be.able-PART(NEG)
“If he dies my Light of the Moon (proper name) will not be able to
inherit the throne” (SK:85). This construction is used to combine
the negative and conditional meanings.
d) the imperative form with zero expression: Tv(IMP) seci
ahån si gisun gisure-rakå o-ci
elder.brother you word speak-PART(NEG) be(AUX)-CONV
mim-be gisure se-ci uju be geks�e
I(bi/min-)-ACC speak(IMP) say(AUX)-CONV head ACC nod(IMP)
(ma. gehes�e-)
“Elder brother (or sir), if you can’t say a word and want to speak,
nod your head” (SK).
This sentence contains two analytical constructions. The first of them,
Tv-rakå oci, is used to denote the negative and conditional meanings
together. The second, Tv(IMP) seci, is used to denote the meanings
of desire and imperative together.
e) the perfect participle in -ha/-he/-ho: Tv-ha/-he/-ho bici
ehe niyalma de hajila-ha bi-ci,
bad people DAT become.intimate.with-PART be(AUX)-CONV
te sain niyalma ojo-rakå bi-he now good man be-PART(NEG) be(AUX)-PART “If (he) was intimate with bad people, he could not be a good man now” (ZAKH:199; QW); donji-mbihe bi-ci, urgun i doro-i hear-PAST be(AUX)-CONV joy GEN rule-GEN acana-me gene-ci aca-mbihe go.to.meet-CONV go-CONV meet(AUX)-PAST urgun i doro “congratulations;” Tv-ci aca- = should, ought, must; “If (I) heard (about your arrival), (I) should come to congratulate (you)” (PASH2:85; QW). In these two sentences the analytical construction, being a predicate of a subordinate clause, denotes an unreal condition which does not correspond with reality. doài-ha (ma. dosi-ha) bi-ci tere sapali mangusi enter-PART be(AUX)-CONV that yellow bull-mangus horgon’i (ma. horho) dorgi de dedu-he bi-he bi cattleshed-GEN inside DAT lie-PART be(AUX)-PART COP
gorel-2.p65
317
4/17/02, 10:59 AM
318
part four
“When (he) had entered that yellow bull-mangus was lying in the cattleshed;” muke de makta-ha bi-ci tere haha jui water DAT throw-PART be(AUX)-CONV that man child hendu-me … say-CONV “When (they) have been thrown into water that fellow says … ;” moro be ungku-me (ma. ungke-) sinda-ha bi-ci cup ACC turn.upside.down put-PART be(AUX)-CONV tere moro dorgi-ci hacinge (ma. hacingga) jete-re this cup inside-ABL all.kinds.of eat-PART omi-re yali buda gemu tuci-he drink-PART meat cooked.cereal all appear-PART “When (he) put down the cup, having turned it upside down, inside this cup all kinds of meals, drinks, meat, and cooked cereal appeared” (SK:84). In the last two examples the analytical construction Tv-ha bici is used to indicate that the main action occurs while the state of the subject, which is the result of completing the dependent action, continues. 5) Constructions where the functional component is expressed by the concessive converb derived from the verbs bi-, o-, se-: Tv-ra bicibe; Tv-ra ocibe; Tv-ha bicibe; Tv-ha ocibe; Tv-ki secibe. These constructions are used to combine the perfect and imperfect meanings of participles or the modal meaning of a desire with the concessive meaning of subordinate predicates. 6) Constructions where the functional component is expressed by the perfect converb: Tv-rakå ofi ere n’alma (ma. niyalma) ere i erdemu be this man this GEN capability ACC sa-rakå o-fi s�ue (ma. s�uwe) uàa-me know-PART(NEG) be(AUX)-CONV totally pull-CONV gama-hai deal.with-CONV “Since this man didn’t know about this capability (here: magic), he pulled (the reins) again and again” (SK). tere wehe be tuk’i-me (ma. tukiye-me) jailabu-re håsun that stone ACC raise-CONV move.aside-PART power
gorel-2.p65
318
4/17/02, 10:59 AM
319
morphology
tesu-rakå o-fi, sele i suffice-PART(NEG) be(AUX)-CONV iron GEN fakài (ma. faksi) jui suhemci (ma. suhecen) be gai-fi wehe workman child small.axe ACC take-CONV stone be håala-me (ma. håwala-) efule-fi ACC split-CONV break-CONV “Since the power was not sufficient to raise the stone and to move it aside, the son of a blacksmith, having taken an axe, split the stone into pieces” (SK:85). The analytical construction Tv-rakå ofi is used to combine the meanings of negation and cause. 7) Constructions where the functional component is expressed by the converb in -nggala/-nggele/-nggolo derived from the verb o-: Tv-ra onggolo jura-ra o-nggolo start-PART be(AUX)-CONV “before starting” (ZAKH:198). 8) Constructions where the functional component is expressed by the dative of the perfect and imperfect participles derived from the verbs bi-, o-, se-: bisirede, bihede, ojorode, ohode, serede, sehede. Here are some patterns of these analytical constructions: a) Tv-me ohode bi ara-me o-ho-de min-i I write-CONV be(AUX)-PART-DAT I(bi/min-)-GEN = my gucu uthai boo-de dosinji-ha friend then house-DAT enter-PART “When I was writing, my friend entered into the house” (ZAKH:150); aikabade sin-i baru bai-me o-ho-de, if you(SG)-GEN toward ask.for-CONV be(AUX)-DAT bisi-re-i teile in-de acabu-me bu-ci, be-PART-GEN only he-DAT please-CONV give-CONV teni inu i gese then correct GEN like Tv-ra/-re/-ro i teile “to the extent of … ;” “If (he) asks you for (smth.) and if (you) give him what (you) have to please him, that will be all right” (PASH2:120; QW); b) Tv-ki serede belkutei daha-ra urse afa-ki PN follow-PART people attack-OPT
gorel-2.p65
319
4/25/02, 1:57 PM
320
part four
se-re-de, belkutei bejihiye-me hendu-me … say(AUX)-PART-DAT PN console-CONV say-CONV “When people, who follow Belkutei, want to attack him, he says: ... ” (PASH:10); sakda ahån i uttu aisila-me wehiye-me old elder.brother/sir GEN like.this help-CONV support-CONV ulin bu-ki se-re-de, ai gelhun akå wealth give-OPT say(AUX)-PART-DAT what timid there.is.not ali-me gai-rakå accept-CONV take-PART(NEG)
gelhun akå “dare to …”, “fearlessly” (NL:104);
“If you, elder brother (sir), wish to give me wealth to help me and
support me, how can I dare not to accept (it)?” (ZAKH:202);
tere han loho be goci-fi saci-ki this khan sword ACC pull.out-CONV hack-OPT se-re-de tere han’i sargan say(AUX)-PART-DAT this khan-GEN wife han’i gala-deri jafa-ha khan-GEN hand-SEP grasp-PART “When khan, having pulled out the sword, wanted to hack (him), the khan’s wife grasped the khan’s hand” (SK:85). The analytical construction Tv-ki serede is used to combine the meaning of desire together with conditional or temporal relationships between main and dependent clauses of a complex sentence. c) Tv-ha/-he/-ho sehede ere-ci sain ningge be sabu-ha se-he-de, this-ABL good SBSTR ACC see-PART say(AUX)-PART-DAT geli tere be sain, se-mbime, ere sain se-rakå again that ACC good call-CONV this good say/call-PART(NEG) kai COP
“You call (thing) good, if you haven’t seen anything better than it;
when you again see something better, you will call it good and what
you called good before you would not call good any longer”
(PASH2:113; QW).
d) Tv-rakå sehede
doose hendu-me damu ardashån banuhån weile-me
Taoist.priest say-CONV only effeminate lazy work-CONV
jobo-ro de mute-rakå ayoo suffer-PART DAT can-PART(NEG) not.likely
gorel-2.p65
320
4/17/02, 10:59 AM
321
morphology
se-he-de, mute-mbi se-me jabu-ha say(AUX)-PART-DAT can-IMPF say-CONV answer-PART “When a Taoist priest said: (I am afraid) only that (you) might be lazy and effeminate, and not capable of doing hard work, (one) answered: (I) can” (PASH2:271; LD). In this sentence the analytical construction Tv-rakå sehede in combination with the word ayoo “not likely” is used to denote the modality of apprehension or “anxious expectation” (see also 5.8.3). Besides, the form of the imperfect participle in the dative (sehede) is used to denote the temporal relationships between principal and dependent clauses. All constructions which have been noted above, include two components. However, in the case when one component that is a form of an autosemantic verb is analytical, the whole construction contains three components, for example: ala-me bi-he bi-ci (o-ci); alaha bi-he bi-ci (o-ci); ala-ha bi-he bi-fi (o-fi); ala-ha bi-he bi-cibe (o-cibe), etc. Note should be taken that the verbs bi- “to be,” “to exist;” o- “to be,” “to become,” and se- “to speak,” “to say” may function not only within analytical constructions where the first component is expressed by an autosemantic verb, but with nouns where these verbs display their primary lexical meaning. Here are some examples: jiha akå o-fi tuttu uda-hakå money there.is.not be-CONV therefore buy-PART(NEG) “Being without money, one didn’t buy (smth.)” (ZAKH:192); aika sain bithe bi-ci, min-de emu udu debtelin be if good book be-CONV I-DAT one several volume ACC juwen bu-fi håla-ki loan give -CONV read-OPT juwen bu- “to lend,” “to give on loan” (NL:168); “If you have a good book (manuscript), please, lend me one or several volumes (of it) to read;” ama-i bisi-re-de gemu tere-i gånin be father-GEN be-PART-DAT all that-GEN thought ACC tuwa-mbi look.at-IMPF “If the father exists, all consider his thoughts (feelings)” (ZAKH:201; QW); ulin akå bi-cibe, jui bi property there.is.not be-CONV son COP “Although there is no property, but there is a son” (ZAKH:205).
gorel-2.p65
321
4/17/02, 10:59 AM
322
part four
5.10. Interrogative Forms of Verbs In Manchu there are a number of interrogative particles which combined with verbal forms, construct interrogative verbal forms, positive and negative. The interrogative particle ni may be written separately or together with the verbal form as one word. It is normally used at the end of a sentence or a period of speech: si ere be aibide baha-fi sa-ha bi-he-ni you(SG) this ACC where can-CONV know-PART be(AUX)-PARTINT “How could you know this?” It should be mentioned that the universal interrogative particle ni is used not only after verbal forms but after nouns as well: ere ai turgun ni this what reason INT “What is the reason for this?” (ZAKH:216). Following negative verbal forms, the interrogative particle ni changes into the element n which is written with the preceding verbal form as one: akå + ni > akån. Here are some examples: si sabu-rakå-n you(SG) see-PART(NEG)-INT “Have not you seen?;” abka gala-rakå-n sky clear.up-PART(NEG)-INT “The sky has not cleared up, has it?;” gene-ki se-rakå-n go-OPT say(AUX)-PART(NEG)-INT “(You) want to go, don’t you?” (ZAKH:216). In the interrogative form the words sain “good,” “well,” yargiyan “true,” “real,” “truth,” “reality” change into the forms saiyån and yargiyån: si saiyån “How are you?;” yargiyån “Is it true?”. The interrogative particle ni may be used to express an exclamation when it is expressed in the form of a question. Intonation plays an important role in the formation of these sentences, however,
gorel-2.p65
322
4/17/02, 10:59 AM
323
morphology
it has never been studied in Manchu. Here are some examples: dule i tuba-de bi-he-ni really he that.place-DAT be-PART-INT “But he really was there?!”; suwem-be hendu-he-ngge, yala uthai you(suwe/suwen- PL)-ACC speak-PART-NR indeed suddenly ji-he-ni come-PART-INT “(We) have been talking about you, and (you) suddenly came!;” dule si uba-de bi-fi niyalma-i baru this.way you(SG) this.place-DAT be-CONV people-GEN toward gisure-mbi ni talk-IMPF INT “This way you have stayed here and talk to people!” (ZAKH:219). Attached to verbal forms, the interrogative particle o also serves to express a question. The following verbal forms may be followed by the interrogative particle o: -ha/-he/-ho + o > -hao/-heo/-hoo; -ra/ -re/-ro + o > -rao/-reo/-roo; -mbi + o > -mbio; bi + o > bio. Following the vowel a, e or o, this element is pronounced as [u], whereas following the vowel i, it sounds as [yu] (Zakharov, 1879:217). Here are some examples: ba-ha-o “Didn’t you get (smth.)?;” yabubu-ci ojo-ro-o put.into.effect-CONV be(AUX)-PART-INT “Is it possible to put into effect?;” ere gese geli kooli bi-o this same also rule COP-INT “Are there any rules of this type?;” tuba-de gene-ki se-mbi-o that.place-DAT go-OPT say(AUX)-IMPF-INT “Do you wish to go there?” (ZAKH:217); sim-be geli niyalma se-ci you(si/sin- SG)-ACC also human.being/man call-CONV o-mbi-o be(AUX)-IMPF-INT “Is it possible to call you a man?” (PASH1:54; QW).
gorel-2.p65
323
4/17/02, 10:59 AM
324
part four
The interrogative particle o may follow not only predicates expressed by verbal forms but also by other parts of speech, for example, nouns or pronouns. This happens in cases when the predicate is expressed by other parts of speech or there is no need to repeat the verbal form, as in the following examples: geren niyalma dorgi fala-ci aca-ra-ngge many people inside be.exiled-CONV meet(AUX)-PART-NR we-be-o who-ACC-INT
Tv-ci aca- = should, ought, must;
“Of all people who must be exiled-CONV, whom?”;
tere niyalma horonggo-o that man powerful-INT “Is that man powerful or not?” (ZAKH:218). To emphasize a question, a combination of the two interrogative particles is often used: ni + o > nio. The interrogative element nio is written separately from the preceding word. It is mostly used in sentences where there are no finite verbal forms: ere aika sin-i sa-ra-ngge nio this something you(si/sin- SG)-GEN know-PART-NR INT “Is it the (same) that you know?”; ere sain akå nio this good COP.NEG (there.is.not) INT “This is not good, is it?” (ZAKH:217). Two more interrogative verbal forms in -mna and -cina may be observed in samples of Sibe dialect recorded by V.V. Radlov and classical Manchu by Iv. Zakharov respectively. It is likely that these forms developed from the interrogative particle na combining with the imperfect and conditional converbs in the following fashion: -me + na > -mna; -ci + na > -cina (Lebedeva & Gorelova, 1994:83; Zakharov, 1879:180-1). Here are some examples from the Sibe dialect: sue (ma. suwe) sain’i bi-mna you(PL) good-GEN be-INT
gorel-2.p65
324
4/17/02, 10:59 AM
325
morphology
“You do live well, don’t you?;”
sargan jui, muse jue (ma. juwe) niyalma be
female child we(INCL) two person ACC
acabu-re arga baga-mna join-PART way can-INT
“Girl, could you (find) a way to join both of us, a son and a mother?”
(SK:83).
In Manchu, a question may be expressed by a specific syntactic
construction, or by repeating verbal words, mostly participles, the
first of them being used in the positive form and the other in the
negative. In some cases, the last participle is followed by one of the
interrogative particles. Thus, the following patterns of interrogative
syntactic constructions can be found: Tv-ra Tv-rakå; Tv-ra Tv-rakån;
Tv-ra Tv-rakå bio; Tv-ra Tv-rakå semeo; Tv-ra Tv-rakå nio; Tv-ha Tv
hakå; Tv-ha Tv-hakån; Tv-ha Tv-hakån biheo; Tv-mbi Tv-rakå; Tv-mbi
Tv-rakån; Tv-mbio Tv-rakån; Tv-mbihe Tv-mbihekå; Tv-mbihe Tv
mbihekån. Here are some examples:
jide-re jide-rakå nio come-PART come-PART(NEG) INT “Will one come or not?” gene-mbi gene-rakå-n go-IMPF go-PART(NEG)-INT “Does one go or not?” songgo-ho songgo-hakå-n cry-PART cry-PART(NEG)-INT “Did one cry or not?” (ZAKH:218); ce kemuni ji-mbi-o jide-rakå-n they as.is.customary come-IMPF-INT come-PART(NEG)-INT “Shall they come as is customary or not?” (ZAKH:217); ài tuta-mna tuta-rakå you(SG) stay.on-INT stay.on-PART(NEG) “Will you stay (any longer) or not?” (SK:83). A question is frequently expressed by the word mujangga “true,” “real,” “correct,” “truly,” “actually,” “indeed” followed by the interrogative particle o: mujanggao/mujanggo “is it true that ... ?”; “can it be that ... ?” This interrogative word normally follows a participle in the negative form:
gorel-2.p65
325
4/17/02, 10:59 AM
326
part four
ojo-rakå mujanggao be-PART(NEG) is.it.true.that?
“Is it true that it is impossible?” (ZAKH:218).
6. Adverbs. Modes of Word Formation. Different Semantic Classes of Adverbs In Manchu adverbs form a grammatical class of words which de note attributes ascribed to a quality or an action. Adverbs are un inflected, i.e. they don’t possess grammatical word forms, but are characterized by the lexical and derivative correlation with all gram matical classes of autonomous words and by the presence of partic ular morphemic means used for forming adverbs. The last charac teristic is optional. The basic syntactic function of adverbs is the attributive one in the broad sense. That is the ability to modify verbs, other adverbs and, finally, the whole sentence. From the point of view of historical morphology, all adverbs are divided into pronominal, nominal, and verbal. The pronominal adverbs are the most archaic ones: uba “here,” tuba “there,” uttu “like this,” “so,” “thus,” tuttu “like that,” “so,” “thus.” By their origin the majority of nominal adverbs are stereotyped case forms having chiefly spatial-temporal semantics such as the dative, the ablative, etc., for example: dade “originally,” “at first,” “in the beginning” (< da “root,” “base,” “foundation” + de = DAT); babade “everywhere” (< ba ba “every place,” “everywhere” + de = DAT; ba “place”). Adverbs may arise from certain verbal forms, mostly converbs: eitereme “thorough ly,” “generally,” “in any case” (< eite- “to do thoroughly” + me = imperfect converb); neneme “formerly,” “previously,” “beforehand” (< nene- “to be first,” “to be ahead” + me = imperfect converb). In accordance with their derivative structure, all Manchu adverbs can be distributed amongst two groups. The first group includes adverbs devoid of symbolic value and the second one comprises derivative (with symbolic value) adverbs. Adverbs belonging to the first group are represented by a small number of words. The majority of these words have lost the correlation with grammatical classes of words. At the same time, many of them, for example, those having semantics of place, are common for all Tungus-Manchu languages. The following adverbs may be considered as non-derived: ne “now,” “at present,” “current;”
gorel-3.p65
326
4/17/02, 12:35 PM
327
morphology
te “now,” “at present;” teni “just,” “then and only then,” “for the first time;” uthai “then,” “thereupon,” “at once,” “immediately;” doigon “beforehand,” “previously;” erde “early,” “early in the morning;” sikse “yesterday;” uba “here;” tuba “there;” tule “outside;” ambula “greatly,” “very much;” asuru “very,” “exceedingly;” kejine “a lot;” umesi “very,” “to a high degree” (umesi and asuru are used only with adjectives, kejine and ambula are used with adjectives as well as verbs); labdu “many,” “much;” komso “few,” “little,” “a little;” majige “a little,” “a little bit;” uttu “so,” “thus,” “like this;” tuttu “so,” “thus,” “like that;” gese “like,” “same;” emgi, sasa “together;” geli “also,” “still,” “again;” esi “certainly,” “of course,” etc. As shown in 2.2, object nouns that have time, place or cause semantics may function in a sentence as adverbial modifiers of time, place or manner. On the grounds that they may act as functional analogues of adverbs, these nouns are often ascribed to this part of speech. They are nouns denoting various notions related to time (seasons, times of day, etc.): inenggi “day,” “in the day-time,” “by day;” dobori “night,” “at night;” yamji “evening,” “in the evening;” cimari “morning,” “tomorrow;” coro “after tomorrow;” niyengniyeri “spring,” “in spring;” juwari “summer,” “in summer;” bolori “autumn,” “fall,” “in autumn;” tuweri “winter,” “in winter,” etc. The following object nouns are related to orientation in space: dergi “top,” “head,” “east,” “upper;” “above,” “over;” fejergi “bottom,” “underneath;” dorgi “the inner part,” “inside;” tulergi “the outer part,” “outside;” amargi “back,” “north,” “behind;” julergi “front,” “south,” “in front of;” cargi “that side,” “opposite,” “beyond;” wargi “west,” “right (side),” “under neath,” “under;” hanci “nearness,” “closeness,” “near,” “close,” “close by;” goro “distance,” “distant,” “far,” “far-away,” “far off,” “a long way off,” “far (from).” Motivated adverbs are characterized by a distinct correlation with other grammatical classes of autonomous words. It is possible to distinguish the following subgroups of derivative adverbs which are formed from nominal and pronominal words: 1) adverbs derived from nominal words by the marker of the gen itive: butui “secretly” (< butu “dark,” “hidden,” “secret” + i = GEN); cihai “as one wishes,” “according to one’s desires” (< ciha “wish,” “desire” + i = GEN); kemuni “often,” “still,” “yet” (< kemun “mea sure,” “model,” “rule,” “regulation” + i = GEN); manggai “merely,” “simply” (< mangga “hard,” “difficult” + i = GEN).
gorel-3.p65
327
4/17/02, 12:35 PM
328
part four
In the structure of some adverbs of this group, one can easily recognize the genitive case marker although nominal words with the same stems but without this marker cannot be found in the language: balai “in vain,” “vainly,” “carelessly,” “falsely,” “unreasonably;” gaitai “suddenly,” “by chance;” uthai “then,” “thereupon,” “immediately,” “at once;” 2) adverbs derived from nouns (including those that have seman tics of place) by the marker of the dative: dade “originally,” “at first,” “in the beginning” (< da “root,” “base,” “foundation” + de = DAT); fejergide “underneath” (< fegergi “underneath” + de = DAT); dergide “above” (< dergi “top,” “above,” “over,” “upper,” “east” + de = DAT); dorgide “inside” (< dorgi “the inner part,” “inner,” “inside” + de = DAT); tulergide “outside” (< tulergi “the outer part,” “outer,” “outside” + de = DAT); 3) pronominal adverbs derived from pronominal words by the marker of the dative: ubade “at this place,” “here” (< uba “here” + de = DAT); tubade “at that place,” “there” (< tuba “there” + de = DAT); aide “where?,” “whither?,” “why?,” “how?” (< ai “what?,” which?,” “how?” + de = DAT); aibade, aibide “where?” (< aiba, aibi “where?” + de = DAT); 4) pronominal adverbs derived from pronominal words by the marker of the ablative: ereci “after this,” “hence,” “from this” (< ere “this” + ci = ABL); tereci “after that,” “then,” “from that” (< tere “that” + ci = ABL); ubaci “hence,” “from here “ (uba “here” + ci = ABL); tubaci “thence,” “from there” (< tuba “there” + ci = ABL); aibaci, aibici “whence?,” “from where?” (< aiba, aibi “where?” + ci = ABL); 5) adverbs derived from nouns, mostly from those that have semantics of time, by the postposition dari “each,” “every:” inenggidari “every day,” “daily;” biyadari “every month,” “monthly;” aniyadari “every year,” “yearly;” mudandari “every time;” 6) quantitative adverbs derived from numerals by the dative: emu de (oci) “in the first place” (< emu “one” + de = DAT); jaide “in the second place” (< jai “second” + de = DAT); 7) other quantitative adverbs are formed by adding the words geri (nggeri) “time,” “number of times;” jergi, mari, mudan “time” and the
gorel-3.p65
328
4/17/02, 12:35 PM
329
morphology
postposition dari “each,” “every” to numerals: emgeri, emu mudan, emu mari “once;” juwenggeri, juwe mudan “twice;” ilanggeri “thrice,” etc.; 8) adverbs derived from nouns by adding the negative particle akå: derakå “without shame,” “without face” (dere “face,” “reputation”); erin akå “never” (erin “time”); fahån akå “without courage” (fahån “courage”); fakjin akå “without support,” “helpless” (fakjin “support”); hercun akå “inattentive,” “not paying attention” (hercun “attention”). Adverbs may be formed from verbal forms: 1) adverbs derived from participles by the marker of the dative: emembihede (< ememu “some” + bihe = perfect participle from the verb bi- “to be,” “to exist” + de = DAT); 2) used permanently in the adverbial function, some imperfect converbs (form in -me) shifted their category status from converbs to adverbs: ambarame “extraordinarily,” “extremely” (ambara- “to do on a large scale”); aname “in order,” “in sequence,” “one after another,” “one by one” (ana- “to push,” “to extend,” “to appoint”); bireme “completely,” “thoroughly,” “universally” (bire- “to rush (into),” “to roll (dough, seed, etc.),” “to attack”); eitereme “generally,” “in any case,” “thoroughly” (eitere- “to do thoroughly,” “to do repeatedly”); dahime “again” (dahi- “to repeat,” “to do again”); hahilame “fast,” “quickly,” “urgently” (hahila- “to act guickly or urgently,” “to hurry”); neneme “formerly,” “previously,” “beforehand” (nene- “to be first,” “to be ahead”); nurhåme “connected,” “in a row,” “in a series” (nurhå“to be connected,” “to be in series”). It appears that the pronominal adverbs adarame “how?,” “why?” and enteheme “eternally,” “always” can be included in this group, however, the verbal stems from which they are formed, are unknown. 3) used permanently in the adverbial function, some perfect converbs (form in -fi) shifted from converbs to adverbs: toktofi “certainly,” “surely,” “without fail” (tokto- “to fix,” “to settle,” “to determine,” “to decide”); 4) used permanently in the adverbial function, some converbs in -tai/-tei/-toi shifted their category status from converbs to the class of adverbs: banitai “by nature,” “inborn” (bani- “to live,” “to be born”); biretei “totally,” “universally,” “completely” (bire- “to rush (into),” “to
gorel-3.p65
329
4/17/02, 12:35 PM
330
part four
roll (dough, seed, etc.),” “to attack”); bucetei “until death,” “scorning death,” “to the last,” “desperately” (buce- “to die”); cohotoi “especially,” “particularly” (coho- “to do especially, “to consider as the most important aspect”); fars�atai “fearlessly,” “recklessly,” “without heed for life” (fars� a- “to risk one’s life,” “to act carelessly”); s� anggatai “finally,” “indeed,” “actually” (s� angga- “to finish,” “to terminate successfully,” “to be accomplished”); waliyatai “to the death,” “without regard for one’s own safety” (waliyata- “to fling about,” “to lose (face),” “to throw around).” The adverb entehetei “forever” presumably has the same origin. 5) negative forms of the imperfect participle very often undergo adverbialization: cikirakå “not snugly fitting” (< ciki- “to insert or attach snugly,” “to fit exactly” + rakå = negative form of the imperfect participle); lakcarakå “uninterruptedly” (< lakca- “to break off” + rakå = negative form of the imperfect participle); baktarakå “extremely” (< bakta- “to contain,” “to encompass” + rakå = negative form of the imperfect participle). Derivative adverbs are formed on the basis of the non-derived ones and nouns that have semantics of place: 1) pronominal adverbs derived by the suffix -si: absi “how?,” “where to?,” “whither?” (aba “where?”); uttusi “in this direction,” “over here” (uttu “like this”); tuttusi “in that direction” (tuttu “like that”); gåwabsi “to another place,” “elsewhere” (< gåwa “other,” “another” + ba “place” + si); 2) the same suffix -si can be seen in the structure of the following adverbs with semantics of place: amasi “backward,” “to the back,” “toward the back;” casi “in that direction,” “thither,” “there” (cala “over there,” “on the other side,” “previously”); desi “upward;” julesi “forward,” “southward;” tulesi “outward,” “toward the outside,” “inside out” (tule “outside”); 3) some adverbs that have semantics of place contain the suffix -ri in their structure: deleri “on top” (dele “top,” “on top”); dolori “inside” (dolo “inside,” “the inside,” “inner”); juleri “in front;” oilori “on the surface,” “on the outside” (oilo “surface,” “outside”). Adverbs with semantics of place seemingly preserved the ancient
gorel-3.p65
330
4/17/02, 12:35 PM
331
morphology
case markers in their structure. These markers are absent both from classical Manchu and the Sibe dialect. The suffix -si ~ -s�i ~ -hi ~ -ki ~ -gi was presumably formerly a marker of the directive, and the suffix -ri < -li was a marker of the prolative. The suffix -ri is, by its origin, a component of the composite suffix -deri (the separative in classical Manchu, and the ablative in Sibe) (Lebedeva & Gorelova, 1994:88). 4) the suffix -hun/-hån is also used to derive adverbs with semantics of place: fusihån “down,” “downward,” “westward;” wasihån “down,” “downward,” “westward;” wesihun “up,” “upward,” “eastward.” The suffix -hun/-hån is typical for nouns with semantics of quality, and it is evidence of what suggests a rather complicated path of development of these adverbs. Presumably, the nouns fusihån, wasihån, wesihun were derived from the adverbial words fusi, wasi and wesi respectively, and then, being used as adverbial modifires of place, they shifted to adverbs. The adverbs wasi “downward” and wesi “upward” have been preserved in the Sibe dialect (Lebedeva & Gorelova, 1994:88). Here is one example: deài waài tua-ci (ma. tuwa-), umai jaka sabu-rakå upward downward look-CONV NEG thing see-PART(NEG) “When having a look upward and downward, (one) saw nothing” (SK:88). It is interesting to note that adverbs can also be derived by adding the words seme or sere, by origin the impefect converb and the imperfect participle, respectively, of the verb se- “to say,” “to mean,” to the onomatopoeic words: bodor seme “mumbling through the teeth” (bodor = the sound of mumbling); bur bur seme “gurgling forth,” “swelling up (of a water spring)” (bur bur = the sound of a water running out of a spring); dar seme “shivering” (dar = the sound of shaking from cold or fright); fiyak fik seme “suddenly,” “with sudden movements,” “without deliberation” (fiyak fik = the sound of jumping from the bush); fer far seme “weakly,” “fluttering slowly like a butterfly in flight” ( fer far = the sound of a butterfly or a bird fluttering); gang gang seme “like wild geese crying” (gang gang = the sound of a flock of wild geese calling); hei hai seme “crying,” “with cry” (hei hai = the sound of crying); hiyor hiyar seme “neighing,” “strongly,” “obstinately” (hiyor hiyar = the sound of a horse neighing); håwasa hisa seme “stepping on
gorel-3.p65
331
4/17/02, 12:35 PM
332
part four
dry leaves” (håwasa hisa = the sound of stepping on dry leaves); je ja seme “screaming,” “shouting loudly” ( je ja = the sound made by men working); kaka kiki seme “laughing,” “with a laugh” (kaka kiki = the sound of happy laughter); kalar seme “kindly,” “courteously,” “harmoniously;” kohong kohong seme “coughing,” “with cough” (kohong kohong = the sound of repeated coughing); konggir seme “ringing,” “tinkling” (konggir = the sound of a small bell); mung mang seme “lowing,” “bellowing,” “roaring” (mung mang = the sound made by cattle or deer); putur seme “drop by drop,” “bit by bit,” “gradually,” “unevenly” (putur = the sound of a large bird taking off); sir siyar, sir sar seme “walking slowly” (sir siyar, sir sar = the sound of grass and leaves moving slightly), etc. This derivational pattern is very productive since the onomatopoeic words are very diverse and widespread in Manchu. From the synchronic point of view, the words seme and sere merely signal the fact that onomatopoeic words are used in the adverbial function. Originally the combination of onomatopoeic words with word forms seme or sere literally meant: speaking like this (that), speaking in this (that) way. Here are examples: ere dobori emu jaka tok tak seme uyunggeri tanta-ha this night one thing ONMT say-CONV nine.times strike-PART “This night one thing struck nine times with the sound “tok tak” (the sound of repeatedly striking a hollow wooden object);” arkan seme jabu-re sagda (ma. sakda) mama i barely say-CONV go-PART old old.woman GEN adali o-hobi like become-PAST “[They] became like an old woman who barely moved” (SK:90). From both the semantic and syntactic points of view all adverbs can be divided into two groups, namely qualitative and non-qualitative (adverbs of circumstance). The qualitative adverbs denote the quality of an attribute or an action. The adverbs of circumstance denote temporal, spatial, causal or final relationships which are specified in such terminological word-combinations as adverbs of time, adverbs of place, adverbs of cause, adverbs of purpose, adverbs of manner, and adverbs of degree. In accordance with their semantics, all adverbs can be divided into the following semantic subgroups:
gorel-3.p65
332
4/17/02, 12:35 PM
333
morphology
1) Adverbs that have semantics of qualitative evaluation. Most specialists include nouns with semantics of quality in this group of adverbs. Some examples of these are: sain “good,” “well;” ehe “bad,” “evil,” “badly;” tob “straight,” “upright,” “right,” “just;” getuken “clear,” “lucid,” “understandable,” “clearly;” arsari “ordinary,” “common,” “everyday,” “commonplace,” “commonly;” enggici “secret,” “secretly,” “privately,” etc. These nouns cannot be considered as adverbs proper although they are frequently used to indicate the quality of an attribute or an action. Here are some more adverbs belonging to this semantic group: acu facu “with loving tenderness;” mekele “in vain,” “vainly,” “emptily,” “merely.” Many qualitative adverbs originated from imperfect converbs. The latter, being permanently used in the function of adverbial modifiers, shifted their category status from converbs to adverbs: dahåme, dasame “again;” dabume “including,” “comprising;” enculeme “separately,” “in addition,” etc. 2) Adverbs that have semantics of quantitative evaluation. This subgroup comprises the following quantitative adverbs: labdu “many,” “much;” komso “few,” “little,” “a little;” ududu “several,” “a number of;” utala “so many (much) as this;” tutala “so many (much) as that;” heni, majige “a little,” “a bit;” daruhai “often,” “continually,” “regularly;” tongga “rare,” “rarely,” “few,” “seldom;” dabali “excessively,” “exceeding,” “too;” dembei “greatly,” “in high degree,” exceedingly;” elemangga “on the other hand,” “on the contrary,” “still more,” “especially;” fulu “excelling,” “surpassing,” “better;” hon “very,” “most,” “too;” jaci “frequently,” “too,” “very;” nememe “moreover,” “on the contrary,” “especially;” umesi “very,” “to a high degree.” Numeral (number) adverbs can also be included in this subgroup. That is adverbs which semantically and formally coincide with several subgroups of numerals and denote the so-called “coefficient of repitition”: emgeri “once,” juwenggeri “twice,” ilanngeri “three times,” etc., as well as collectivity: juwenofi (juwe nofi) “two persons,” “two people,” “the two of (us, you, them),” “both (of);” ilanofi “the three of (us, you, them),” etc. (see Section 3 on Numerals). Those derived from numerals by words mari, mudan “time (as in one time, two times, etc.),” dari “each,” “every” (postposition) also belong to this group of adverbs: biyadari “every month,” “monthly;” inenggidari “every day,” “daily;” mudandari “every time;” emu mari “once,” “one time;” juwe mari “two times;” ilan mari “three times,” etc.
gorel-3.p65
333
4/17/02, 12:35 PM
334
part four
3) Adverbs that have semantics of time: ne “now,” “at present” “currently;” te “now,” “at present;” dade “at first,” “in the beginning,” “originally;” neneme “formerly,” “previously,” “beforehand;” amala “after,” “later;” aifini “a long time before,” “much earlier;” jakan “just now,” “not long (in duration),” “recently;” emembihede “sometimes,” “now and then,” “at times;” erde “early,” “early in the morning;” enenggi “today;” coro “the day after tomorrow;” sikse “yesterday;” cananggi “the day before yesterday,” “previously;” doigonde “before hand,” “previously;” uthai “at once,” “immediately;” amasi “after,” “henceforth;” erileme “at the right time,” “on time,” “from time to time.” 4) Adverbs that have semantics of place. Three subgroups can be distinguished amongst adverbs of place semantics. Those belonging to the first subgroup are used to indicate the location of an object with non-motion verbs, for example: eri “this is here;” uba “here;” tuba “there,” tule “outside;” dele “on top;” fejile “under,” “underneath;” juleri “in front;” amala “behind;” bakcin “opposite;” babade “everywhere.” Adverbs belonging to the second subgroup are used to indicate the place from where an object moves. These words may be defined as pronominal adverbs with deictic meaning: ubaci “from here;” tubaci “from there;” aibaci, aibici “from where.” Adverbs belonging to the third subgroup are used to indicate the place where an object moves to: ubade “to this place,” “here;” tubade “to that place,” “there;” absi “to where;” gåwabsi “to another place,” “elsewhere”12. The fourth subgroup includes adverbs indicating the place along or through which an object moves: alirame “along a mountain;” bigarame “through the wilds,” “across the wilderness;” butereme “along the mountain’s foot;” dalbarame “along the side;” dalirame “along the shore,” “along a riverbank;” golorome “through other provinces,” “across other provinces.”
7. Interjections and Onomatopoeic Words Interjections take special place amongst other grammatical classes of autonomous words. One group of interjections serve to express emotions. The following interjections are used to denote affection, 12 Norman in his Lexicon gives other translations of the forms “ubade” and “tubade,” i.e. as “at this place” and “at that place” respectively (NL:290, 281).
gorel-3.p65
334
4/17/02, 12:35 PM
335
morphology
surprise or praise: ai, adada ebebe = an exclamation of surprise; adage = an exclamation of affection used when patting an old person or a child on the back; ajaja = an exclamation of affection; aya = an interjection of praise or surprise. The interjections a, ak, ok are used to express fear or fright. The interjections a, ara, ara fara, are serve to express pain, grief or regret; ake is used when touching something hot. The interjection ei is a signal of derision. Another group of interjections is used as signals of volitional inducements. Thus, the interjections ei, oi are used to attract attention. The interjection ma is used when handing something to another person. The interjection of response is a, and a a is an interjection of casual response. The interjections en, en en seme, je are used to answer affirmatively. The exclamations gåi gåi, gåje are used to call a falcon, and wer wer is used to call a dog. The interjection a si is used for driving chickens or birds, and cu is used to set a dog on someone or thing. The latter also means “get out!” The interjections takasu and takålu mean “wait a moment,” “just a moment.” A number of interjections can reveal different meanings depending on what sensations are caused by a certain action. Thus, according to Zakharov, the interjection a may express fear, as well as response or affirmation; the interjection ai is used to express pain, as well as surprise; the interjection aya is a signal of surprise, as well as fear or pain; the interjection e is used as an exclamation of affirmation or response; the interjection ei serves to express fear or for calling people’s attention (Zakharov, 1879: 315). According to Norman, the interjection a serves to express fear or response, e is used as an exclamation of surprise or exasperation, as well as for getting someone’s attention; the interjection ei is used for getting attention or as a signal of derision (NL:3, 69, 71). It is important to note that all interjections are characterized by syntactic isolation and absence of formal links witn preceding and following elements in the spoken chain. Sometimes it is difficult to distinguish interjections from onomatopoeic words. Thus, Zakharov classified the language elements ei ei and kaka kiki as interjections used to express laughter, while Norman defines these entities as onomatopoeic words: ei ei = a sound of a derisive laughter, and kaka kiki = a sound of happy laughter (Zakharov, 1879:315; NL:71, 170). Undoubtedly, these language entities are related to the expression of laughter. The onomatopoeic words are ascribed to the interjections by
gorel-3.p65
335
4/17/02, 12:35 PM
336
part four
tradition because of lack of precise creteria for distinguishing them, although it seems that in Manchu most onomatopoeic words can be clearly distinguished from interjections on semantic grounds. Onomatopoeic words are extremely widespread in Manchu, and a great number of such words had been fixed by the time Zakharov’s Lexicon was published. Many of them can be also found in Norman’s Lexicon. An impressive quantity of significant language material is presented in the article published by G. Stary. He has also described several important patterns according to which new verbs had been derived from onomatopoeic words in Manchu (Stary, 1981:209-22). Onomatopoeic words express attitudes of human beings and all living creatures caused by certain actions or states as well as states of inanimate objects which are in motion. Though their semantics is rather vague, they express a certain sort of propositional content. They refer to the situation (the state of affairs) as a whole when the subject and the predicate do not constitute separate parts of the utterance. In his Grammar, Zakharov distinguished 29 subgroups of most frequently used onomatopoeic words (Zakharov, 1879:315-20). According to him, the onomatopoeic words may express the following types of sounds caused by different actions of all living creatures including human beings as well as by moving of inanimate objects (with meanings as given in Norman’s Lexicon): 1) cough: keng = the sound of coughing; keng kang = the sound of many people coughing or clearing their throats; kohong kohong, korkong korkong = the sound of repeated coughing; 2) vomiting: ok, or; 3) snoring: ko ka = the sound of sniffing or snoring; kor = the sound made when something gets caught in the throat; hir har = the sound of wheezing in the throat or snoring; kår kar = the sound made when there is phlegm in the throat; 4) spitting and clearing one’s throat: hak = the sound of clearing one’s throat; pei = the sound of spitting; pei pai = the sound of repeated spitting; piyas pis = the sound of spitting to show contempt; 5) breathing and sighing: s�uwang s�ang = the sound of sighing and groaning; ha = the sound made by breathing on frozen objects; he fa = the sound of gasping, panting; ho ha = the sound of sighing made by a tired person; pu = the sound of blowing with the mouth; pu pur = the sound made when blowing out the fire;
gorel-3.p65
336
4/17/02, 12:35 PM
337
morphology
6) crying, weeping or moaning: hei hai = the sound of crying; u u = the sound of weeping; gar gar = the sound made by a small baby; miyang = the sound of a child’s crying; miyang ming = the sound of many children crying; miyar mir = the sound of children crying; miyar miyar = the sound made by a baby crying; 7) talking, muttering, murmuring: badar = the sound of speaking without due deliberation, speaking wildly; bar bar, bar bir, ger, gur, lor = the sound of many people tallking together; dur = the sound of many people talking or laughing; 8) shouting of different nature: ang = a scream used in battle; ar = the sound of calling or shouting; gar = the sound of shouting; gar gar = the sound made by a small baby; gar gir = the sound made by a group of people arguing; ger gar, tek tak = the sound of shouting and quarreling; ja ji = the sound made by many people screaming; jar, je ja = the sound made by men working; jor = the sound of many humans screaming; jor jar = the sound of people screaming; kå ca = the sound of fighting; kåwak cak = the sound of fighting with poles or sticks; 9) being squeamish: ek = the sound made when one is squeamish; 10) stammering, stuttering: keke kaka; 11) chuckling or snorting: for for, hiyor hiyar = the sound made by a horse snorting or neighing; hor = the sound of neighing, whinnying; kus = the sound made when one is not able to keep from laughing; pus = the sound of bursting out laughing; sur suk = the sound of chuckling; 12) noise and rustle made by walking: bir biyar = the sound made by many people going in a body (according to Norman, bir biyar imitates the sound made by hanging to the floor (of clothing); NL:30); håwasa hisa = the sound made by stepping on dry leaves; håwar hir = the sound of clothing rubbing together; kata kiti = the sound of shoes treading on a hard surface; kutu kata = the sound on walking feet; kafur = the sound of walking on ice or snow (crunching); kafur kifur = the sound made in stepping on ice or snow; kåwas kis = the sound made by someone dragging his feet; picik pacak = the sound of walking in mud; 13) shivering, shaking from cold or fright: dar; dordon dardan; 14) shouting or roaring of animals and barking of dogs: ang ang = the sound made by camels and donkeys; fer fer = the sound of yelping dogs; ger = the sound made by snarling dogs; giyang = the sound of a dog barking; gur = the sound of snarling, growling; jor = the sound
gorel-3.p65
337
4/17/02, 12:35 PM
338
part four
of many dogs, chickens or animals screaming; miyang ming, miyar miyar = the sound made by the young of deer, roe and sheep; mung mang = the sound made by cattle or deer; 15) singing, chirrupping, screaming of birds, crowing of roosters: fer far = the sound of a butterfly or a bird fluttering; gang gang = the sound of a flock of wild geese calling; gar gir = the sound made by a flock of crows; giyar giyar, giyar gir = the sound made by monkeys and birds; giyong = the sound made by the wings of a phoenix in flight; goko = the sound of chickens cackling; gon gan = the cry of a goose or swan; gåli gali = the call of the oreole; gåwar gåwar = the cry of ducks, frogs or doves; ja ja = the sound made by a bird when it is caught; jak jik = the sound of birds screaming in flight; jang jing = the sound made by birds looking for one another; jar jir = the sound made by birds early in the morning; jing yang = the sound of birds singing harmoniously; jor jar = the sound of birds screaming; kiyar = the sound made by a wild hawk; kiyar kir = the cry of alarm made by birds of prey and martens; kior = the sound of a bird taking off suddenly; kotor = the sound of pheasants flying or taking off; kotor katar = the sound of a flock of pheasants flying; per par = the sound of birds beating their wings; potor patar = the sound of a group of birds flying; pur = the sound of birds taking flight; putur, putur potar = the sound of a large bird taking off; yur yar = the sound of birds flying in row (in file); 16) sounding of various insects: jar, jar jar, jir jir = the sounds made by crickets; jar jir, jor jar = the sound of grasshoppers and other insects flying; per par = the sound of insects beating their wings; per pir = the sound of a grasshopper taking off; piyang = the sound of crickets chirping; sar sar = the sound of grasshoper flying, the sound of insects chirping; yang ing = the sound of insects flying; 17) jingling, ringing, tinkling of bells, drums, etc.: cang cang, cang cing, kanggir kinggir, tang tang = the sound of bells; jing jiyang = the sound of flutes and stone bells; hanggir hinggir = the tinkle of bracelets and anclets; holor halar = the sound of many bells; kalang, kalang kiling = the sound of metal or stone objects banging together; kalar kalar = the sound of metal objects hitting one another; kalar kilir = the sound of keys or small bells jingling; kanggir = the sound of metal or porclain falling; kiyalang = the sound of a single bell; the sound of metal colliding with another object; konggir = the sound of a small bell; konggir kanggir = the sound of many small bells; kung cang, kåwang cang = the sound of drums and cymbals; kutur = the sound of incessant drumbeating; s�ak s�ik = the sound of tingling; tang ting = the sound
gorel-3.p65
338
4/17/02, 12:35 PM
339
morphology
of hitting iron; tong tong = the sound of a shaman’s drum; tung tang = the sound of bells and drums; tung tung = the sound of a drum; ung = the sound of a bell; yang yang = the sound of bells ringing; 18) beating, chopping, hitting, striking: cing cang = the sound of chipping wood and ice; giyok = the sound of crashing; hob, kåwas = the sound of chopping wood; kåwata kiti = the sound of a hard object striking something; piyak = the sound of slapping; pok = the sound of hitting something; s�uwak = the sound of hitting with a whip; s�uwak s�ik = the sound of several whips striking; tak = the sound made by hitting something solid; tak tik = the sound made when chopping wood; tang ting the sound of chopping a tree; tok = the sound of striking a hollow wooden object; tok tok = the sound of repeatedly striking a hollow wooden object; yak = the sound made by a whip; 19) flaming, blazing; firing muskets; exploding firecrackers; breaking dried wood; mowing with a sickle, tearing clothing, etc.: cur = the sound of a rocket flying; cur car = the sound of firecrackers and rockets; hång = the sound of a fire; hår, hår hår = the sound of blazing, flaming; kes = the sound made by a knife, sharp, shear when cutting off; kes kes = the sound of breaking dried wood or brushwood; kifur kiyafur = the sound of crunching or gnashing; kiyak = the sound of dried wood breaking; kiyak kik = the sound of a large tree splitting; kåwang = the sound of knocking on wood; the sound of explosion; kåwas kis = the sound of a breaking stick; the sound of a sickle mowing; pak = the sound of exploding firecrackers; pak pik = the sound of many small firecrackers going off; pes = the sound of something soft being torn or broken; pes pas = the sound of something soft ripping; pus pos = the sound of silk or leather tearing; tes = the sound of rope, thread, or a leather thong breaking under stress; tur tar = the sound of a muskets firing; 20) breaking, crunching, gritting: fasar, feser = the sound of breaking something into many pieces; kafur kifur = the sound of crunching on teeth; kakår = the sound of gritting the teeth; katur = the sound of crunching; katur kitur = the sound of eating hard brittle things (like ice); kete kata = the sound of dried fruits and biscuits eaten by children; kiyas = the sound of something brittle breaking; kiyatur kitur = the sound of clods being crushed under wagon wheels; piyas = the sound of breaking, splitting; 21) noise made by something falling on the earth, on the floor, in water: fak fik, pata piti = the sound of fruit falling from a tree; fatak = the sound of falling objects; håwalar hilir = the sound of fishnets
gorel-3.p65
339
4/17/02, 12:35 PM
340
part four
in water; kanggir = the sound of metal or pocelain falling; kanggår = the sound of a wall falling; kanggår kinggår = the sound of a large structure collapsing; katak kitik = the sound of an object falling from a high place; kung, kutung = the sound of a large object falling to the ground; pak = the sound made by a hard object falling on the floor; sar sar, sar sir = the sound of leaves falling; 22) noise made by wind: hoo, hoo hio = the sound of the wind blowing violently; håwasar = the sound made by desiccated plants in the wind; s�eo s�a = the sound made by the wind blowing; s�or s�ar = the sound of a storm, the sound of wind and rain; 23) noise made by a thing being dragged, or drawn: fasak = the sound made by a beast or a bird suddenly emerging from a thicket; fosok = the sound of a wild beast leaping from cover; hoso hasa = the sound of shaking paper; håwar = the sound of a thing being dragged on the ground; kåwas kis = the sound of dragging sacks of grain on a floor; s�uwar = the sound of a sword being drawn from a scabbard, a snake moving rapidly, or an arrow passing through the air; 24) noise made by water: bur bur = the sound of water gurgling forth, of a water spring swelling out; busu busu = the sound of a rain; cor = the sound of water gushing forth uninterruptedly; fotor, jir jir = the sound of water bubbling; hång hiyong = the sound of the tide; håwalar = the sound of wading in water, the sound of flowing water, the sound of splashing; håwanggar = the sound when it is raining copiously; håwanggar håwalar = the sound of water roaring and splashing; konggor = the sound of pouring water; pocok, pocong = the sound of things hitting the water; tab, tab tib = the sound of dripping water; yur yur = the sound of rippling water when flowing; 25) noise made by running horses, carts, or people: hång hiong = the sound of running horses; kakång kikung = the sound made by a heavily loaded cart or by a heavy load; ketek katak = the sound of cart wheels on a rough surface; kiyakång, kiyakång kikung = the sound made by a heavily loaded wagon, with a rumbling sound; kiyatur kitur = the sound of clods being crushed under wagon wheels; kunggur = the sound made by empty wagons; 26) piercing some object throughout: cos = the sound of ricocheting or rebounding; kiyob, kob = the sound of arrow piercing a target; pos = the sound of an arrow piercing an object; pus = the sound of piercing something; tos = the sound made by an arrow or other like implement piercing some object cleanly; 27) the sound of thunder: kunggur = the sound of heavy thunder;
gorel-3.p65
340
4/17/02, 12:35 PM
341
morphology
kunggur kanggar = the sound of thunder; kiyatar = the sound of roaring of thunder; 28) snapping at each other: kab = the sound of snapping, beating of a pack of dogs; kab kib = snapping at each other of dogs fighting or beating; 29) the sound of something sizzling when frying: car cir. Many of the above listed onomatopoeic words are polysemantic to a high degree, for example: the word ha may imitate the sounds 1) made by breathing on frozen objects, 2) made when eating something hot or salty, 3) the cry of a bird of prey when it sees a man; the word combination hång hiyong may imitate the sound: 1) of the tide, 2) of running horses; the word jar may imitate the sounds 1) made by men working hard, 2) made by crickets, 3) made by a bone-headed arrow; the word combination kåwas kis imitates the sounds 1) made by someone dragging his feet, 2) of a sickle mowing, 3) of dragging sacks of grain on a floor, 4) of a breaking stick (NL:120, 140, 156, 183). As one can see from the above material, most interjections are monosyllabic, while most onomatopoeic words are disyllabic with the second syllable closed, ending in the sounds b, k (a letter rendering this sound is k with two dots), ng, r or s. Many onomatopoeic words are formed by repeating the same word (or element): cang cang (the sound of bells); kohong kohong (the sound of repeated coughing); kalar kalar (the sound of metal objects hitting one another); sar sar (the sound of grasshoppers flying). The others are formed by adding the second element which includes vowels opposed to those of the first one. In other words, if the first word (element) contains back vowels the second has front (or neutral) vowels and vice versa. Two elements may have different vowels, for example: her har (the sound of clearing one’s throat); je ja (the sound made by men working); keng kang (the sound of many people coughing or clearing their throats); kaka kiki (the sound of happy laughter); kunggur kanggar (the sound of thunder); pak pik (the sound of many small firecrackers going off); tab tib (the sound of dripping water); s�or s�ar (the sound of a storm, also of wind and rain). As mentioned in Section 6, adverbs may be formed by adding the word form seme (originally the imperfect converb derived from the verb se- “to say”) to an onomatopoeic word, for example: the onomatopoeic word ar imitates the sound of calling or shouting, and the adverb ar seme means “shouting loudly;” the onomatopoeic word
gorel-3.p65
341
4/17/02, 12:35 PM
342
part four
cang imitates the sound of bells, and the adverb cang seme means “hard,” “fast;” the onomatopoeic word gar imitates the sound of shouting, and the adverb gar seme means “loudly;” the onomatopoeic word hiyor hiyar imitates the sound of a horse neighing, and the adverbs hiyor hiyar seme means “neighing,” as well as “strongly,” “obstinately;” the onomatopoeic word tur imitates the sound of a galloping horse, and the adverb tur seme means “at a gallop,” “fast.” This mode of adverbial formation is mostly used when there are no adverbs proper with the same meaning in the language. In the absence of a governing verb, the converb seme becomes the main verb, and the verbal stem se- may assume any grammatical form required: pak pik = the sound of many small firecrackers going off; pak pik se-he “(a rocket) went off;” for for = the sound of neighing or snorting; for for se-he “(a horse) got snorted;” hår seme “blazing,” “flaming;” hår se-he “became a bit tipsy”13. In Manchu, combinations of onomatopoeic words with the verbal stem se- became the basis for forming new verbs: ek se- “to feel sick,” “to be tired of,” “to be annoyed” (ek = the sound made when one is squeamish); gåli gali se- “to cry like an oriole” (gåli gali = the sound of calling an oriole”); hiyor hiyor se- “to be robust” (hiyor = the sound of the feathers on a flying arrow; hiyor hiyar = the sound of a horse neighing or snorting; hiyor hiyar seme “strongly,” “obstinately”); kaka kiki se- “to chuckle,” “to giggle,” “snigger” (kaka kiki = the sound of happy laughter); kek se- “to be pleased,” “to be gratified,” “to be refreshed” (< kek; kek seme). In a number of such combinations, the morpheme se started to be pronounced as one with the onomatopoeic word: cibse- “to be quiet” (cib cab “quiet,” “still;” cib cib seme “very quiet,” cib seme “quietly”); kekse- < kek se- “to be pleased,” “to be gratified,” “to be refreshed;” kiyalang se- “to ring,” “to clang” (kiyalang = the sound of a single bell, the sound of metal colliding with another object). Changed to the morpheme si, the morpheme se gave rise to a new derivational suffix si: caksi- “rattle,” “to vibrate,” “to beat a cymbal” < cak cak se- < cak seme; koksi- 1) “to cackle,” 2) “to cry (of pheasants in spring time)” < koko se- < koko = the sound made by chickens; tuksi- “to pound,” “to throb (of the heart)” < tuk seme “pounding (of the heart)” (Zakharov, 1879:321; NL:41, 178, 283; Stary, 1981:213, 219). 13
gorel-3.p65
Norman translates ek se- as “to be tired of,” “to be annoyed with” (NL:72).
342
4/17/02, 12:35 PM
343
morphology
G. Stary gives some other patterns of formation of new verbs from onomatopoeic words or adverbs corresponding to them. According to his language material, new verbs may be formed by the suffixes -(r)a/-(r)e, -(r)i, -(r)u, -(r)s�a/-(r)s�e, -ja/-je, -da/-ta: dakdars�a- “to jump (upon),” “to act in aggressive manner” < dakda dakda / dakda dikdi “in great leaps;” giltars�a- “to shine brightly,” “to gleam” < gilta gilta “shining,” “glowing”/ gilta gilti “glittering,” “gleaming” (see also giltarila- “to shine,” “to glitter”); sebers�e- “to drip” < seb sab = the sound made by falling drops; seb sab aga- “to rain in scattered drops;” sab “dripping, “in scattered drops;” caru- “to fry,” “to deep-fry” < car seme “painful (of a burn),” “at a breath,” “at a stroke” < car cir = the sound of meat sizzling (used to describe a painful wound); mura 1) “to roar,” “to low,” “to bellow,” “to neigh,” 2) “to call deer with a whistle” < mung mang seme “lowing,” bellowing,” “roaring” < mung mang / mung mung = the sound made by cattle or deer (see also murada“to roar from rage;” murakå “a whistle for calling deer;” muran “a battue held at the time of the deer-breeding season;” mur mar “obstinate,” “stubborn”); guluje- “to slitter” < gulur seme “slittering;” gåwari- “to croak (of frogs)” < gåwar se- “to quack,” “to croak” < gåwar gåwar = the cry of ducks, frogs or doves; miyasida- / miyasita- “to walk unsteadily,” “to fly unsteadily (of an arrow)” < miyasi miyasi “unsteady,” “staggering;” miyasihida- / miyasihita- “to walk unsteadily” < miyasihi miyasihi “unsteady,” “staggering” (Stary, 1981: 209-22). These verbs and some others formed by the suffixes -s�a/-s�e, -ja can also be found in Norman’s Lexicon, for a example: heos� e - “to be hesitant,” “to be undecided” < heo seme “passable,” “sufficient but not perfect,” “fairly well;” cakja- “to become crusted over (of snow)” < cak seme 1) “neatly,” “firmly (wrapped),” 2) “freezing” (NL:40, 128). Some new verbs appeared as a result of combining verbs and adverbs, which derived from the onomatopoeic words by means of the word form seme: cing seme da- “to flame” (cing seme “flaming,” “flashing”); dar seme s�urge- “to shake from cold or fright” (dar seme “shivering”); fir seme ara- “to write fluently” (fir seme “quietly, “calmly,” “fluently”), heo seme banji- “to get by fairly well” (heo seme “fairly well”), etc. (NL:45, 54, 88, 128).
gorel-3.p65
343
4/17/02, 12:35 PM
344
part four 8. Syntactic Words
Syntactic words are not used as parts of the sentence. They serve to convey various semantic-syntactic relationships between autonomous words, word-combinations, clauses or sentences. The main function of syntactic words is connecting, distinguishing or correlating words of basic parts of speech or clauses ( or sentences). Syntactic words play a very important role in the Manchu language, much more noticeably than they do in other languages of the Tungus-Manchu language community. Together with case forms or alone, they denote relationships between words within wordcombinations. In combination with case markers of the preceding nouns, they are used to denote additional grammatical meanings within word-combinations. They may also link clauses within complex sentences or simple sentences within compound ones. According to their grammatical meaning and functions, all syntactic words can be distributed amongst several classes, namely, postpositions, conjunctions, syntactic words of a conjunctional type and particles. Although most syntactic words can be clearly included in one class or another, in some borderline cases their distinctions can be blurred. Amongst syntactic words there are ones which may convey relationships between autonomous words within wordcombinations as well as those between clauses within complex sentences or simple sentences within compound ones. All syntactic words occupy the position of a governing word within a word-combination or a sentence. According to the general rule of word order, subordinate words precede those by which they are governed, syntactic words, with a few exceptions, normally stand after the words or syntactic units which they govern, and directly before those on which they depend. The link between syntactic and autonomous words or syntactic units is normally expressed only by the position they take. Autonomous words may be governed by syntactic ones by means of certain case markers. The case markers are placed between the autonomous and syntactic words. It should be stressed that the sphere of syntactic words in the Manchu language requires further investigation. 8.1. Postpositions In Manchu, postpositions are uninflected syntactic words which serve to denote syntactic relationships between nouns or a noun and a verb.
gorel-3.p65
344
4/17/02, 12:35 PM
345
morphology
Postpositions govern nouns, pronouns or nominalized (or substantivized) participles which they follow. If governed words occur in certain case forms, postpositions follow the latter specifying their syntactic meanings. Manchu postpositions are represented by two structural types. Basic (non-derived) postpositions or those the etymology of which is not clear, are ascribed to the first structural type: baru “to,” “toward;” emgi “with,” “together with;” gese “like;” jalin “because of,” “on account of;” manggi “toward,” “with;” sasa “with,” “together with;” tule “outside,” etc.14. Derivative postpositions can be ascribed to the second structural type. Most derivative postpositions are of nominal origin: dade “in addition to” (< da “base,” “foundation,” “root,” “source” + de = DAT); funde “in place of,” “instead of,” “for” (< fun + de = DAT); dorgide “in inside” (< dorgi “the inner part,” “inner,” “inside” + de = DAT); jakade “to the presence of,” “up to,” “by,” “in front of” (< jaka “thing,” “object” + de = DAT); ninggude “on top of” (ninggu “top,” “on top” + de = DAT); sidende “in between” (< siden “space,” “interval,” “a while” + de = DAT); songkoi “according to,” “in accordance with” (< songko “trace,” “track” + de = DAT); tulergide “on the outside,” “outside” (< tulergi “the outside part,” “outer,” “outside” + de = DAT). Derivative postpositions may also be characterized by converbal origin. Several forms, mostly of the imperfect and terminal converbs, via adverbs, shifted to the class of postpositions: dahame “according to” (< daha- “to follow” + me = imperfect converb); isitala “up to,” “until” (< isi- “to reach,” “to arrive,” “to come to” + tala = terminal converb); otolo “until” (< o- “to become,” “to be” + tolo = terminal converb); tuwame “in accordance with,” “depending on” (< tuwa- “to look,” “to look at” + me = imperfect converb). Each postposition governs a certain noun case, and according to that, all of them can be distributed amongst the following groups: 1) postpositions requiring the nominative case: jaka “together with;” 2) postpositions requiring the genitive: adali “like;” baru “toward;” 14 The postpositions gese “like,” emgi, sasa “with” were defined before as adverbs: gese “like,” “same;” emgi, sasa “together” (for details see Section 6). In Manchu, words frequently shift from one grammatical class to another, partly losing their original lexical meaning and developing specific grammatical meanings instead.
gorel-3.p65
345
4/17/02, 12:35 PM
346
part four
dele, deleri “on top;” dergi “above,” “over;” dorgide “inside;” emgi “with;”
gese “like;” fejergi, fejile “under’, “underneath;” funde “instead of,” “in
place of,” “for;” hanci “near,” “by;” jakade “in the presence of,” “up
to,” “by,” “in front of;” jalin “on account of;” manggi “toward,” “with;”
sasa “with;” sidende “in between;” songkoi “according to,” “in
accordance with;” teile “only;”
3) postpositions requiring the dative: isitala “up to,” “as soon as;”
otolo “up to;”
4) postpositions requiring the ablative: amala “behind;” amasi “toward
the back;” ebsi “up till now;” julesi “toward the front;” tulgiyen
“besides,” “otherwise.”
Postpositions which are characterized by verbal origin, require the
same case that underlying verbal stems do. Thus, the postpositions
dahame “according to” (daha- “to follow”) and tuwame “in accordance
with,” “depending on” (tuwa- “to look,” “to look at”) require the form
of the accusative; aisilame “with the help of” (aisila- “to help”) and
acame “in accordance with” (aca- “to meet,” “to join”) require the
form of the dative. Some postpositions may govern two case forms.
Here are some examples:
ere baita sin-ci tulgiyen, gåwa niyalma ainaha seme
this matte you(si/sin- SG)-ABL apart other people surely
inu sa-rkå
so know-PART(NEG)
ainaha seme “surely,” “categorically,” “without fail” (NL:9);
“As to this matter, nobody (literally: other people) is as knowledgeable
in it as you” (PASH2:80; QW);
bi hono age i jakade gene-fi, tacibu-re be I still sir GEN to go-CONV instruct-PART ACC donji-ki se-me gåni-re bade listen-OPT say(AUX)-CONV intend-PART since
sin-i boo-de feliye-rakå
you(si/sin- SG)-GEN house-DAT walk/take.steps-PART(NEG)
mujangga-o
actually-INT
mujanggao= is it true that …?; can it be that …? (NL:203);
“Since, for some time, I have had intentions to come to you to listen
gorel-3.p65
346
4/17/02, 12:35 PM
347
morphology
to your instructions, sir, how I cannot help stepping into your house
(and sitting there for a while)?” (PASH2:73; QW);
uttu o-ho-de, bi teni sin-i
like.this be(AUX)-PART-DAT I just you(si/sin- SG)-GEN
funde gene-fi fas�s�a-ci o-mbi
instead.of go-CONV exert.efforts-CONV become(AUX)-IMPF
uttu ohode “if so;”
fas�s�a-ci o- “can exert efforts;”
“If so, I can go instead of you and do my best (to solve your problem)”
(PASH2:107; QW).
In accordance with their semantics, all postpositions can be divided
into several subgroups: 1) locative postpositions; 2) temporal
postpositions; 3) comitative postpositions,
4) causal postpositions, 5) postpositions of similarity and 6) restrictive
postpositions.
1) The group of locative postpositions includes the following ones:
baru “toward;” dalbade “on one side;” dele, deleri “on top;” dergi “above,”
“over;” dergide “on;” dorgide “inside;” fejile, fejergi “under,”
“underneath;” hanci “closed to;” isitala “up to;” jakade “by,” “up to,”
“in front of;” juleri “in front;” julesi “toward,” “toward the front;”
julergi “in front of;” ninggude “on top of;” sidende “in between;” tule,
tulergi “outside;” tulergide “outside.” Here are several examples of
locative postpositions. All of them govern the genitive with the
exception of isitala “up to” which governs the dative:
muse te boo i baru muda-ki we now house GEN toward come.back-OPT “We wish to come back home now;” ere mama i baru gene-he this old.woman GEN to go-PART “(Someone) went to this old woman;” ejen muduri baru hendu-me … khan dragon to say-CONV “The khan says to the dragon …;” ere sagda uce i jakade te-he this old.man door GEN by sit-PART “This old man sat by the door;”
gorel-3.p65
347
4/17/02, 12:35 PM
348
part four
ere sunja niyalma i jakade iàina-ha (ma. isina-ha)
this five person GEN up.to reach-PART
“[They] reached these five men;”
uhuri (ma. uheri) da-ci dube-de iàitala gemu …
altogether beginning-ABL end-DAT to all
“(When) all from the beginning to the end...”(SK:91).
One can notice that a number of locative postpositions are materially
identical to nouns which have semantics of place: amala “behind;”
amargi “back,” “behind;” amasi “backward,” “to the back,” “toward
the back;” dele “top,” “on top;” deleri “top,” “surface,” “superficial;”
dergi “top,” “above,” “over,” “upper;” dorgi “the inner part,” “inner,”
“inside;” fejergi “bottom,” “under,” “underneath;” julergi “front,” “in
front of,” “before;” juleri “front,” “in front;” ninggu “top,” “on top;”
tule, tulergi “the outer part,” “outer,” “outside.” Most of these nouns
may reveal characteristics of a noun, adjective or adverb when
functioning in a sentence. Some of them are regularly followed by
the dative: dergi de, dorgi de, ninggu de, tulergi de. Frequently used to
modify qualities and actions, they assumed, through the function of
an adverbial modifier, the ability to specify locative relationships
between categorematic words. These forms did not totally lose their
relationships with their autonomous counterparts, but they became
partly desemantized in the course of time. Their nominative function,
that of naming objects, transformed into a syntactic one, i.e. denoting
syntactic relationships between words of full meaning.
Used as locative postpositions, they are often presented by the form of one of the locative cases, the dative or the ablative. Nevertheless, they are mostly used in the dative which is governed by the main verb. In its turn, the dative form of the syntactic word governs the genitive form of the subordinate noun. Here are several examples of syntactic words which have full counterparts: sourin’i dergi de emu s�ain meihe haya-fi throne-GEN top DAT one white snake coil.up-CONV dedu-hebi lie-PAST “One white snake was lying coiled up on the throne” (compare the syntactic word dergi de “on” with the noun with place semantics dergi “top” in the dative case ); hotun’i (ma. hoton) tulergi de tata-fi settlement(GEN) the.outer.part DAT camp-CONV
gorel-3.p65
348
4/17/02, 12:35 PM
349
morphology
“They camped outside the settlement” (compare the syntactic word tulergi de “outside” with the noun of place semantics tulergi “the outer part” in the dative); mengun’i (ma. menggun) fengseku i cihe (ma. cai) be sagda silver-GEN bowl GEN tea ACC old.man i julergi àinda-fi (ma. sinda-fi) GEN front put-CONV “Put the silver bowl in front of the old man” (compare the syntactic word julergi “in front of” with the noun of place semantics julergi “front”); hotun’i (ma. hoton) hanci iàina-ha (ma. isina-ha) town-GEN near come-PART “(They) came close to the town” (compare the syntactic word hanci “close to” with the name of place semantics hanci “proximity,” “close,” “near,” “near by”) (SK:93). The locative postpositions are significantly more numerous than the others, such as temporal or causal. The explanation is that locative relationships are typical for a simple sentence, within wordcombinations, between nouns or a noun and a verb. Temporal or causal postpositions are more typical for complex sentences, when the relationships between clauses become more actual than those between words. Therefore most syntactic words with temporal or causal semantics occur mainly in complex sentences. Locative postpositions can develop temporal semantics when used in complex sentences. Within complex sentences they show a tendency to shift their category status from postpositions to syntactic words of the conjunctional type. 2) The temporal postpositions are amari “after,” “afterwards;” isitala “up to,” “until,” “as soon as;” manggi “after;” otolo “until.” They are more often used in complex sentences revealing a tendency to shift to the grammatical class of conjunctions (or syntactic words of the conjunctional type). Some postpositions may denote both locative and temporal relationships: amasi “to the back,” “toward the back;” “after,” “henceforth.” 3) The comitative postpositions emgi and sasa “with” govern the genitive case of a noun (or a pronoun). Here are some examples on comitative postpositions:
gorel-3.p65
349
4/17/02, 12:35 PM
350
part four
bi àin-i (ma. sin-i) emgi gene-mbi I you(ài/àin- SG)-GEN with go-IMPF “I shall go with you;” ninggun sargan jui sain sejen de te-fi ele ulha jaka six woman child good cart DAT sit-CONV all cattle thing hacin aha nehu i sasa gemu tuci-ha various slave slave.girl GEN with all leave-PART “Six daughters sat in good carts and left with all the cattle, various property, and slaves” (SK:92). When the word sasa does not occur with the genitive case, it should not be considered as a postposition but as an adverb: àin-i (ma. sin-i) emgi jabu-fi buce-ci banji-ci you(ài/àin- SG)-GEN with go-CONV die-COND live-COND sasa together “(I) will go with you, we shall die or live together”(SK:92). 4) The postposition jalinde “for the sake of” governs the genitive case of a noun: àin’i (ma. sin-i) emu niyalma i jalinde nadan you(ài/àin- SG)-GEN one person GEN for.the.sake.of seven niyalma be wa-ra person ACC kill-PART “For the sake of you, one man, I killed seven men.” Used as a postposition, the word jakade may reveal the same meaning: min’i ama i jakade miau niru-re
I(bi/min-)-GEN father GEN for.the.sake.of oratory paint-PART
jalin ài abka gene
in.order.to you heaven go(IMP)
“You should go to heaven in order to paint the oratory for the sake
of my father” (SK:92).
4) The postpositions with the meaning of similarity are adali and gese
“like.” They govern the genitive case of a noun or follow participles
which occur in the out-of-case form. Here are several examples:
ài buce-ci bi geli buce-he adali kai you(SG) die-CONV I also die-PART like COP “If you die I also shall be like dead;”
gorel-3.p65
350
4/17/02, 12:35 PM
351
morphology sagda mama i adali o-hobi old old.woman GEN like become-PAST “(They) became like an old woman” (SK:92).
6) The restrictive postposition teile “only,” shifted from the grammatical class of adverbs, governs the genitive case of a noun or a participle. Following a participle, the postposition teile means “to the extent of:” jabduhai teile “as time permits,” “as opportunity allows;” muterei teile “with all one can,” “with all one’s capabilities,” “to the extent of one’s power” (NL:152, 206). Sometimes it is used together with the adverb damu which also has a restrictive meaning. Both having restrictive semantics, any of these two words can be omitted, but the restrictive meaning does not vanish: damu ilan sargan’i teile ilan boo de te-he only three woman-GEN only three house DAT live-PART bi COP “Only three women lived in three houses” (SK:92). As mentioned, several postpositions regularly follow participles in out-of-case form. Thus, the syntactic word manggi “after” is often used after the perfect participle or imperative; the postpositions that have the meaning of similarity adali, gese “like,” “same” may also be used after participles; the restrictive postposition teile “only” may govern the genitive case of the perfect and imperfect participles. The syntactic word jakade may also occur after the imperfect participle, and in this case it has the meaning “since,” “when” and reveals a tendency for shifting to the class of conjunctions. It is important to mention, that when occurring after participles, which may serve as predicates of clauses, postpositions change their grammatical meaning to a certain extent. Their main function transforms into that of connecting clauses, not words, and therefore they can rightfully be considered conjunctions (or syntactic words of the conjunctional type) rather than postpositions. One can see that some syntactic words may serve as postpositions or conjunctions depending on what syntactic context they are used in. 8.2. Conjunctions The question of the existence of conjunctions as a separate part of speech in Manchu deserves special consideration. Although almost
gorel-3.p65
351
4/17/02, 12:35 PM
352
part four
all Manchu grammars contain a chapter dedicated to conjunctions, as a grammatical category, this fact is partly explained by the need of translating Manchu texts into a number of European languages. Presumably, the absence of conjunctions was a characteristic of the ancient state of the Manchu language in which the system of participles and converbs was highly developed. Since many syntactic relations might be expressed by participial and converbal forms, there was no need for a large number of conjunctions. The most recent state of the language is characterized by a more developed category of the conjunction which is closely connected with establishing the system of the expanded and the complex sentence. Nevertheless, most syntactic words which can be analyzed as conjunctions, are in a state of transition into this class of syntactic words, and they display their affiliation with other parts of speech at the same time. Besides conjunctions proper, in Manchu there are a number of categories of syntactic words which are used to link syntactic units. These categories include postpositions which can be engaged in the system of complex sentences and syntactic words of the conjunctional type which partly preserve a lexical meaning together with that of linking clauses or sentences, i.e. a syntactic function. Used in the function of linking clauses, i.e. as conjunctions, postpositions normally change their meanings. It is accepted by the majority of manchurologists to categorize most words listed below as conjunctions. In order to classify a word as a conjunction proper, the standard theoretical requirements should be met, however, only a few of these words strictly satisfy these requirements. All Manchu conjunctions can be divided into basic (non-derived) and derivative. From the point of view of morphological structure the conjunctions damu “but,” eici “or,” geli “and,” jai “and,” uthai “then,” “thereupon,” “and then” are morphologically simple, however, these words originally belonged to other parts of speech. They became conjunctions by means of conversion. Thus, the words damu “only;” geli “also,” “stiill,” “again;” jai “again,” “still,” “more,” “later” can be treated as adverbs, the word uthai originally was a pronominal adverb with the meaning “then,” “at once;” the word eici “perhaps” is a parenthetical word. Like many words in Manchu, these ones are also characterized by polysemantics and polyfunc tionality. Depending on the syntactic context they may realize any of their grammatical features and play different roles in a sentence.
gorel-3.p65
352
4/17/02, 12:35 PM
353
morphology
The derivative conjunctions include aikabade “if,” “in the case that;” erinde, fonde “when;” bici, oci “if;” bicibe, ocibe, gojime “though,” “although;” ofi “because;” tuttu oci “if like that’, “if thus,” “in that case;” tuttu ofi “therefore,” “so,” etc. Derivative conjunctions may have nominal, pronominal, adverbial or verbal origin: 1) the following derivative conjunctions are formed from nouns by the dative case: bade “when,” “in the case that,” “if,” “since” (ba “place” + de = DAT); dade “besides (that),” “when” (da “base,” “foundation,” “root,” “source” + de = DAT); erinde “at the time when” (erin “time” + de = DAT); fonde “at the time when” (fon “time” + de = DAT); jakade “when,” “since” (jaka “thing,” “object” +de = DAT); turgunde “because,” “since” (turgun “reason” + de = DAT). The words dade, jakade were classified as postpositions earlier (see Section 8.1). In fact, under certain conditions, they can be used to link clauses within complex sentences, which is the grammatical function of conjunctions or syntactic words of the conjunctional type (see Section 8.3). What is important to stress is that used as conjunctions (or syntactic words of the conjunctional type), they change their meanings. 2) the following conjunctions are derived from pronouns ai “what” and ere “this:” ereni “by this,” “through this,” “from this,” “therefore;” they are often used in a sentence in pairs: aibe ... aibe ... “as ... as ...;” ereni ... ereni ... “hence,” “therefore” (Pashkov, 1963:43). 3) the following conjunctions originated from adverbs including pronominal ones: uttu, tuttu “so;” damu “but” (as an adverb, this word has the meaning “only”); 4) the following conjunctions originated from converbs, mostly imperfect, perfect, and conditional ones, which are derived from the verbs bi- “to be,” “to exist,” o- “to become,” “to be” in their auxiliary function: bime “and” (the form of the imperfect converb of the verb bi-); ofi “because” (the form of the perfect converb of the verb o-); bici, oci, “if” (the form of the conditional converb of the verbs beand o- respectively); bicibe, ocibe “though,” “although” (the form of the concessive converb of the verbs bi- and o-). Several conjunctions originated from converbal forms derived from the verb se- “to say,” “to mean” when it functions in its partly desemantized variant: seme
gorel-3.p65
353
4/17/02, 12:35 PM
354
part four
“that,” “in order to,” “though,” “although;” secibe “though,” “although” (forms of the imperfect and concessive converbs of the verb se- respectively); seci “if” (the form of the conditional converb of the verb se-). Serving as a conjunction, the form of the imperfect converb dahame “as,” “since” being derived from the verb daha- “to follow,” always requires the marker of the accusative be. Seemingly, almost all these forms, perhaps with the exception of the words bime and ofi, cannot be considered as conjunctions proper. Earlier (see Section 5.9), it was already shown that the so-called analytical constructions became the basis on which most conjunctions of converbal origin are produced. Functioning outside of analytical conctructions, these forms partly retain their verbal nature. 5) The following conjunctions originate from participles derived from the verbs bi- “to be,” “to exist,” o- “to be,” “to become,” and se“to say” in their auxiliary function.These participles are mostly used in the dative: bisirede “when,” “at that time, when”(the form of the imperfect participle of the verb bi- + de = DAT); bihede “when,” “if” (the form of the perfect participle of the verb bi- + de = DAT); ojorode “at that time, when,” “when” (the form of the imperfect participle of the verb o- + de = DAT); ohode “when,” “if” (the form of the perfect participle of the verb o- + de = DAT); serede “when” (the form of the imperfect participle of the verb se- + de = DAT); sehede “when,” “if” (the form of the perfect participle of the verb se- + de = DAT). Analytical constructions can also be considered as the basis on which these conjunctions or, to be more precise, syntactic words of the conjunctional type, are produced. Used to connect clauses, these syntactic elements preserve their aspectual-temporal meanings within analytical constructions which they are included in (for details see Section 5.9). Composite conjunctions consist of two elements which are either both simple conjunctions or a simple conjunction and another word belonging to some other part of speech. These include aikabade “if,” “in the case that” (aika = conjunction “if,” bade = conjunction “when,” “in the case that,” “if,” “since”); aika oci “if” (aika = conjunction “if,” “whether,” oci = a syntactic word of the conjunctional type “if”); hono bade “when still” (hono “still,” “yet,” bade = conjunction “when,” “in the case that,” “if”); sere anggala “instead of,” “not only (but also)” (sere = imperfect participle of the verb se-, anggala = postposition “in
gorel-3.p65
354
4/17/02, 12:35 PM
355
morphology
place of,” “instead of,” “rather than,” “not only”); teile akå “not only” (teile = postposition “only,” akå = negative copula/particle); tere anggala, tere dade, terei dade “in addition to” (tere “that,” anggala = postposition “in place of,” “instead of,” “rather than,” “not only;” terei < tere “that” + i = genitive, dade = conjunction “when,” “besides that”); tere onggolo “before that,” “beforehand” (tere “that,” onggolo “before,” “in front”); terei amala “after that” (tere “that” + i = genitive, amala = postposition “after,” “behind”); udu bicibe “although there is,” “though” (udu “although,” bicibe = the form of the concessive converb from the verb bi-); uttu ofi, tuttu ofi “so” (uttu = pronominal adverb “like this,” tuttu = pronominal adverb “like that,” ofi = the form of the perfect converb of the verb o-); uttu ohode, tuttu ohode (uttu = pronominal adverb “like this,” tuttu = pronominal adverb “like that,” ohode = the dative form of the perfect participle of the verb o-). Composite conditional and concessive conjunctions are normally divided by a clause or a syntactic unit into two elements. The first of them is placed at the beginning of the text which has conditional or concessive meaning, and the second element stands at the end of this text. Thus, the text with conditional or concessive meaning (conditional or concessive clauses) is delimited by two syntactic words which serve as parenthesis. Here are some examples: sin-de aika sain bithe bi-ci, min-de you(si/sin- SG)-DAT if good book be-CONV I(bi/min-)-DAT emu udu debtelin be juwen bu-fi håla-ki one several volume ACC loan give-CONV read-OPT juwen bu- “to give a loan,” “to lend;” “If you have a good book, lend me, please, one or two volumes (of it) to read” (ZAKH:201). A clause with conditional meaning, is delimited by the conjunction aika “if” and the form of the conditional converb bici which has not totally lost its lexical meaning of “to be.” aikabade giyangna-me urebu-me tuwancihiya-me if explain-CONV practice-CONV correct-CONV getukele-rakå elucidate-PART(NEG)
o-ci, arbun mudan tongki jijun i durun
be(AUX)-CONV shape sound dot stroke GEN rule
tuta-ra gojime, tere-i jurgan elei survive-PART although that-GEN meaning almost
gorel-3.p65
355
4/17/02, 12:35 PM
356
part four
burubu-re de isina-mbi disappear-PART DAT achive-IMPF “If there is no correct explanation and elucidation (of six classes of word stems—L.G.), then even if the shape (of letters), their fonation, dots, lines, and the rules (of their combining) were maintained, (we) would achieve only vague understanding of their meanings” (PASH2:319; QW). In this sentence the conditional clause is delimited by the conjunction aikabade “if” and the analytical construction getukelerakå oci where the functional element oci is also used to render the conditional meaning. honin deberen udu amtangga bi-cibe, sheep the.young.of.animals although tasty be-CONV geren i angga de acabu-ra-ngge mangga all GEN mourh DAT please-PART-NR difficult
“Although a lamb is tasty, it is difficult to please everybody”
(PASH:61);
udu gai-ki se-me inu baha-rakå although take-OPT say(AUX)-CONV so get-PART(NEG) kai MDL.PTL “No matter how (you) wanted to take (smth.), you will not get (this item)” (PASH2:127; QW). Clauses with concessive meaning are delimited by the conjunction udu “although” and the converbal forms bicibe and seme which also have a concessive meaning. According to their meaning, all Manchu conjunctions are divided into coordinating and subordinating ones. Coordinating conjunctions can be subdivided into 1) copulative (connective), 2) adversative, 3) disjunctive. They are used to connect elements which are syn tactically symmetrical, i.e. which perform identical syntactic functions. These three types of conjunctions differ as the types of logical and/ or semantic relationships they denote. 1) The main function of the copulative conjunctions is to connect elements which are semantically symmetrical. The words geli “also,” “still,” “again;” jai “still,” again;” hono (hono bade) “still,” “yet;” kemuni “still,” “yet” are normally used in the function of the copulative conjunction. The imperfect converb bime “being,” derived from the verb bi- in its auxiliary function, is also used in the role of the copulative conjunction. The conjunction dade “besides (that)” and
gorel-3.p65
356
4/17/02, 12:35 PM
357
morphology
the syntactic word-combinations tere dade, tere anggala “in addition to” also have connective meaning. The syntactic word-combinations sere anggala and teile akå have the connective meaning “not only (but also). Used as a conjunction, the word uthai reveals the connective meaning “then,” “thereupon,” “and then.” 2) The main function of the adversative conjunctions is to connect such syntactic elements which are in opposition. These conjunctions are: bai “but,” akåci “otherwise,” bimbime “however.” 3) The disjunctive conjunctions are used to denote relationships between syntactic elements that indicate the possibility to choose between them. The disjunctive conjunctions are eici “or;” eici ... eici ... “either ... or;” embici “or.” 4) There is another type of conjunctions, i.e. explicative ones, which connect such syntactic elements when the second of them is introduced to explain or concretize the first: duibuleci “for example” (originally the conditional converb from the verb duibule- “to compare,” “to give as an example”); tebici “namely” ( -cina. According to
Zakharov, this form denotes a polite request in the manner of a
question, and it is considered a form of the imperative mood
(Zakharov, 1879:180-1). In the Sibe dialect there is a special
interrogative verbal form in -mna which was formed on the basis
of the imperfect converb (form in -me) and the particle na: -me +
na > -mna. In the course of time it has lost its converbal character
and shifted from converbs to a verb proper, serving as the predicate
of an interrogative sentence:
sue (ma. suwe) sain’i bi-mna?
you (SG) good-GEN be-INT
“Do you live well?” (SK:83).
The interrogative particle o combines with participles and verbs
proper:
damu niyalma be koro ara-fi, beye de tusa ara-ki onle person ACC offence do-CONV self DAT benefit do-OPT se-ci o-mbi-o? say(AUX)-CONV be(AUX)-IMPF-INT
“Is it possible to benefit doing harm to others?” (PASH2:119; QW);
kice-rakå-ci o-mbi-o?
try-PART(NEG)-CONV be(AUX)-IMPF-INT
kicerakåci < kice-rakå + o-ci;
“Is it impossible to try (to do that)?” (PASH2:48; MB);
àin-i (ma. sin-i) jergi gasha be geli n’alma (ma. niyalma)
you(ài/àin- SG)-GEN kind bird ACC like human.being
o-fi eigen gai-ci o-mbi-o?
become-CONV husband take-CONV become-IMPF-INT
“ May I, a human being, take a bird like you as a husband?” (SK:98).
The interrogative particle o may be attached to copulae:
eime-re kooli bi-o?
loathe-PART custom COP-INT
“Why should I loathe you?” (literally: “Is there a custom to loathe
you?”) (PASH2:74; QW);
gorel-3.p65
376
4/17/02, 12:35 PM
377
morphology
geren niyalma-i dorgi fala-ci aca-ra-ngge crowd people-GEN inside punish-CONV meet(AUX)-PART-NR
we-be-o?
who-COP-INT
fala-ci aca- “should be punished;”
“Who is that person (from this crowd) who must be punished?”
(PASH2:15; QW).
The interrogative particle ni is mostly used at the end of a sentence
or a syntactic period to denote a question. Following verbs or
participles, this particle is mostly written separately, however, they
may be occasionally written as one word as well. Here are some
examples with the interrogative particle ni:
ere ai turgun ni? this what reason INT “What is the reason of this?” (PASH2:17; QW); ere gese kooli geli bi-ni? this like norm also COP-INT “Are there any laws similar to this one? (PASH2:46; MB); aina-hai baha-mbi-ni? do.what?-CONV = how get-IMPF-INT “How could it happen?” (PASH2:34; OB); si ere be aibide baha-fi sa-ha bi-he-ni? you this ACC where be.able know-PART be(AUX)-PART-INT “Where could you get to know this?” (ZAKH:216); ere ai bi-he ni? this what be-PART INT “What is this?” ai baita bi-he-ni? what matter be-PART-INT “What is the matter?” (SK:98). The interrogative particle ni is often used after the negative particle akå: we ya de emu hes�u has�u baita akå ni? who what DAT one trifling matter PTL.NEG INT
hes�u has�u “trifling,” “petty,” “annoying” (NL:129);
“Is there anybody who has no trifling duties?” (PASH2:93; QW);
bi ainu taka-rakå ni?
gorel-3.p65
377
4/17/02, 12:35 PM
378
part four
I why know-PART(NEG) INT “Why am I not aware (of this)?” (PASH2:60; JB). In negative participles which are formed by the negation akå, the particle ni often undergoes transformation into n: abka gala-rakå-n ? sky clear.up-PART(NEG)-INT “The sky is clearing up, isn’t it?;” gene-ki se-rakå-n? go-OPT say(AUX)-PART(NEG)-INT “Do you wish to go or don’t you?” (ZAKH:216); tere-i gånin be tuwa-ci o-mbi-o that-GEN thought ACC see-CONV be(AUX)-IMPF-INT ojo-rakå-n ? be(AUX)-PART(NEG)-INT
tuwa-ci o- “possible to see (know);”
“Is it possible to know his opinion (thoughts) about this matter or is
it not possible?” (PASH2:96; QW).
The words sain “good,” “well” and yargiyan “true,” “real,” “truth,”
“reality” have the interrogative forms saiyån “How are you?” and
yargiyån “Is it true?” Here are two examples:
si saiyån “How are you?” (NL:231);
ere baita yargiyun
this event real(INT)
“Is this event true?” (ZAKH:217).
The interrogative particle ni can be used to express an exclamation
in such sentences where the exclamation is combined with a question:
dule i tubade bi-he-ni in.fact he there be-PART-INT/EXCLM “In fact, he was there!” dule si ubade bi-fi niyalma-i baru gurure-mbi in.fact you(SG) here be-CONV people-GEN toward talk-IMPF ni INT/EXCLM “In fact, you are sitting here and talking to people!” (ZAKH:219). To emphasize the question, a combination of interrogative particles is often used:
gorel-3.p65
378
4/17/02, 12:35 PM
379
morphology ere sain akå ni-o? this good PTL.NEG (there.is.not) INT-INT “Isn’t it a good (deed)?” or “Isn’t it good?” (PASH2:18; QW).
4) In old Manchu the following exclamatory particles were widely used: na, ne, no, nu, jiya, jiye, and ya. Even at the time when Iv. Zakharov was writing his Grammar, these particles were not frequently used (Zakharov, 1879:219). Here are some examples: bi-na be-EXCLM “There is!” wakala-rakå na blame-PART(NEG) EXCLM “(You) don’t blame (smb.)!” ere sain jiya this good EXCLM “This is good!” muse uthai yabu-me jiye we(INCL) at.once go-CONV EXCLM “So, we go at once!” absi sain jiye how good EXCLM “How good it is!” inu ya yes EXCLM “correct!,” “exactly!” (ZAKH:219). Particles emphasizing the meaning of words they follow can also be found in Manchu. The particle tome “each,” “every” is presumably a form of the imperfect converb derived from the verbal stem “to-” (compare with ton “number”) which had in the past the meaning “to count” but had gone out of use. The particle tome is used to designate the totality of objects of the same nature, and at the same time it is used to indicate any single object which this totality contains. The particle tome is used as a postposition, occurring only with nouns: ilan niyalma tome sargan gai-ha bi three man every wife take-PART COP “Each of three man has chosen a wife (for himself)” (SK:100). The particle dari “each,” “every” occurs only with nouns which denote periods of time: biyadari “every month,” aniyadari “every year,”
gorel-3.p65
379
4/17/02, 12:35 PM
380
part four
inenggidari “every day.” Since it had lost its morphological independence, the particle dari developed into a form similar to a derivational suffix: muke inengdari (ma. inenggidari) sain’i eye-mbi water every.day good-GEN run-IMPF “Water runs well every day” (SK:100).
10. Reduplication of Words. Grammatical Meanings of Pair Words Applicable to certain morphological classes of words, reduplication and pair words, as grammatical methods, are widely used in Manchu. Both methods are used to form new words, synthetic and analytical forms, phraseological units. Nouns with semantics of place and time are reduplicated and combined in pair words much more rarely than other classes of nominal words. Verbs proper as well as the perfect, conditional, and concessive converbs are never reduplicated or combined in pair words. The imperfect converbs are combined in pair words more frequently than other grammatical classes of words of full meaning. Reduplication is a repetition of autonomous words without changing their phonetic and morphological shape. The pair word is a combination of two synonyms or two words belonging to the same lexical set of words. A pair word may be represented by combination of two antonyms. Being an element of reduplicative or pair word combinations, each word preserves its independent phonic compositon. Both reduplicative and pair word combinations may serve to express plurals (or more properly collectives): jalan jalan “generations,” “worlds” (jalan “generation,” “age,” “world”). Here is one more example: ere enduri hutu ala-ha this spirit desembodied.spirit tell-PART “These spirits have told [something]” (SK:101). The pair word enduri hutu consists of the word enduri “God,” “spirit” and the word hutu “devil,” “disembodied spirit.” These two words belong to the same lexical set, and they are used in this way to indicate plurality.
gorel-3.p65
380
4/17/02, 12:35 PM
381
morphology
adun ulha ulin jaka jetere omire ele hacin herd domestic.animal property thing meal drink all various baitalan gemu bi thing.in.daily.use everything COP “Domestic animals, property, meals and drinks, various things in daily use, everything there is” (SK:101). In this sentence there are three pair word combinations. The first of them, adun ulha consists of two words which belong to the same lexical set: adun “herd” and ulha “livestock,” “domestic animal.” This pair word serves to indicate plurality. The second pair word combination ulin jaka contains two words referring to the same lexical set of words, the word ulin “goods,” “possessions,” “property,” “wealth” and the word jaka “thing.” This pair word combination is also used to express plurality. The third pair word combination jetere omire contains two imperfect participles in their substantive use, that is the participle jete-re “what is eaten,” “meal” ( jete- “to eat”) and the participle omi-re “what is drunk,” “drink” (omi- “to drink”). This pair word is used to express a general notion “subsistence of every kind,” “meals and drinks,” which is associated with the notion of plurality. The plurality of persons or objects, where each person ot object is regarded as separate is expressed by this grammatical device and is used together with the category of plural. Here is an example: emken emken sargan ju-se de ulin nadan jaka hacin one one woman child(jui)-PL DAT property seven thing various ulha morin n’alma (ma. niyalma) tome bu-hei domestic.animal horse people every give-CONV “(He) gave property, various things, domestic animals, horses, people, everything to every (his) seven daughters, one by one” (SK:102). The lexical meaning “every,” “one by one” is expressed twice in this sentence. Firstly, it is expressed by the reduplication of the word emken “one” (emken emken “one by one”) and secondly, by the particle tome “every” placed at the end of the sentence. Double expression of the same grammatical meaning by two different grammatical devices occurs frequently in Manchu texts. Reduplicative and pair word combinations are often used as grammatical devices to indicate grammatical meanings of high degree
gorel-3.p65
381
4/17/02, 12:35 PM
382
part four
of intensity of an action and high degree of qualitative characteristics which a person or an object possesses. High degree of intensity of an action is usually expressed by pair words represented by two imperfect converbs which have synonymous meaning: gemu songo-me (ma. songgo-me) fa-me men’i men’i all cry-CONV be.tired-CONV each/every bou (ma. boo) i baru muda-ha house GEN to return-PART songgome fa- “to weep bitterly (when parting)” (NL:248); “All have returned to their houses weeping bitterly” (SK:102). In this sentence the high intensity of the action is expressed by the pair of the imperfect converbs which may be ascribed to the same lexical set: songgome fame “weeping bitterly” (songgo- “to cry,” fa- “to be tired”). kata-me kangka-me (ma. kanggka-me) buce-he bi dry-CONV be.thirsty-CONV die-PART COP “Suffering from thirst very much, (he) has died” (SK:102). In this sentence the high intensity of the action is expressed by the pair of imperfect converbs which possess synonymous semantics, that is katame kanggkame “suffering from thirst very much” (kata- “to dry” and kanggka- “to be thirsty”). To express a high degree of qualitative characteristics which a person or an object possesses, normally nouns that have the meaning of an object, quality, personal nouns (much more rarely than other classes of nouns) and almost all groups of pronouns are used. Here are some examples: emu amba yafan kåaran (ma. kåwaran) sabu-mbi one big garden courtyard see-IMPF “(He) has seen a big and beautiful garden” (SK:102). The pair word combination yafan kåwaran “beautiful garden” consists of the word yafan “garden,” borrowed from Chinese, and the word kåwaran “courtyard.” ai ai jaka be ambula ba-ha what what thing ACC many get-PART
gorel-3.p65
382
4/17/02, 12:35 PM
383
morphology “(They) have got many various things” (SK:102).
The reduplication of the pronoun ai “what” in the attributive function has the meaning “various” (“what, what”). The permanent use of any reduplicative and pair word combinations has led to a situation where some of them have lost their original semantics. On this basis many idiomatic wordcombinations were formed: giyan giyan i “in detail,” “in proper order” (gian “reason,” “principle,” “order”); emken emken “one by one,” “little by little” (emken “one”); yaya “every,” “each,” “any” (ya “which?,” “what?”); ududu “several,” “a number of,” “many” (udu “how much?,” “how many?”). Reduplication and pair words normally occur in languages which do not belong to the Tungus-Manchu language community, for instance, in Chinese. Seemingly, in Manchu, reduplication and pair words, as grammatical methods, were borrowed from the Chinese language.
gorel-3.p65
383
4/17/02, 12:35 PM
This page intentionally left blank
385
morphology
PART FIVE
SYNTAX
gorel-3.p65
385
4/17/02, 12:35 PM
This page intentionally left blank
387
morphology
PART FIVE
SYNTAX 1. Preliminaries Syntax is usually defined as a linguistic discipline which studies such language entities as word-combinations and sentences, both sim ple and composite. The rules according to which word-combinations and sentences are built from word forms are often also included in syntax. I accept, in this book, a view of the subject and role of syn tax according to which syntax investigates, on the one hand, the rules for linking words and word-combinations, and, on the other hand, the language units within which these rules are realized (Beloshap kova, 1977:5; Cheremisina & Kolosova, 1987:6). The major task of syntax is the study of certain syntactic patterns and rules, according to which connected speech, as a free combination of lexical units, is carried out. A sentence, which serves to form and express human thought, is the central and fundamental unit of syntax. The main distinctive fea ture of a sentence as opposed to language units of lower level is that the sentence is an utterance; it is communicative. This means, first ly, that the sentence reflects a certain situation (an event or a state of affairs) of the outside world, and, secondly, that it possesses the communicative purpose of affirming or negating smth., of asking about smth., or impelling smb. to do smth. According to this, sen tences may be affirmative, negative, interrogative or imperative. As a unit of communication, the sentence realizes its communicative purposes through the syntactic categories of modality and tense. The latter are expressed by the verbal forms of mood and tense, as well as by intonation, modal words, and words denoting localization in time. According to their syntactic structure, sentences are very diverse. Nevertheless, they are mostly realized by word-combinations of varying degrees of complexity. Realized through a word-combination, a sentence possesses predicative structure. This means that it has either a predicative word form, including a nominal (non-ver-
gorel-3.p65
387
4/17/02, 12:35 PM
388
part five
bal) predicate, or two strictly correlated principal parts of a sentence—the subject and the predicate. In reality, these constructions become sentences due to the intonation with which they are pro nounced. Words and non-predicative word-combinations may be come sentences only in certain syntactic conditions including dia logue. When they become sentences, such word-combinations acquire a certain communicative function, a connection to a certain situa tion of the outside world, and a certain intonation (Maslov, 1987:1689). The second aspect of syntactic objects is related to the concept of syntactic relationships. Without syntactic relationships there are no syntactic objects since the latter are always understood as con structive, formed according to a certain pattern. Opposed to each other, the notions of coordination and subordination are the fun damental concepts in the study which investigates syntactic relation ships. Subordination is defined as an inequitable relationship when one component depends on another (or one component is governed by another). The subordinating relationship is mostly used in the structure of word-combinations and it may be: 1) attributive, 2) ob jective, or 3) adverbial. At the level of a sentence, subordination man ifests itself in the relationship between dependent and principal claus es in complex sentences as well as between the principal part of an expanded sentence and the participial or adverbial constructions which it includes. The ways of expressing subordinate relationships at the level of a complex sentence differ from those between words in word-combinations or in a simple sentence. To express subordi nating relationships in the Tungus-Manchu languages the following grammatical means are used mostly: 1) government, 2) agreement, and 3) juxtaposition. Another means of subordination referred to as “reflection” in scientific literature is also found in Manchu. The term “reflection” was initially introduced by V.A. Avrorin to designate a special means of expressing syntactic relationships in attributive wordcombinations (Avrorin, 1960:31-42). Later this term was applied to the attributive word-combinations studied by O.A. Konstantinova and E.P. Lebedeva in the Evenki language (Konstantinova & Leb edeva, 1953:241-2). The term was also used by B.D. Kolesnikova in her “Syntax of the Evenki Language.” She referred to “reflec tion” in Evenki as a special means of expressing syntactic relation ships when subordinating relationships between the principal and dependent components of a construction (word-combination) man
gorel-3.p65
388
4/17/02, 12:35 PM
389
syntax
ifest themselves by means of the form of the principal (governing) component. There are two types of “reflection:” 1) the “reflection” of the person and the number of the dependent component of a wordcombination by means of personal-possessives suffixes which are attached to the principal component; 2) the “reflection” of the quan titative characteristics of the dependent component of a word-combination by means of the grammatical form of the principal com ponent. The first type of “reflection” semantically denotes possessive relationships in the broad sense. The dependent component (deter minatum) of an attributive word-combination formed by means of “reflection,” is normally expressed by a noun, or a personal pronoun of the third person, singular or plural, or a word of another parts of speech which can function in the role of a noun. The principal component (determinandum) of such an attributive word-combination has the possessive suffix for the third person, singular or plu ral, in its structure (Kolesnikova, 1966:37-42). Here are some ex amples in Evenki: Nungan am§k§ abdun-m§-n baka-ra-n. he bear lair-ACC-3.POSS.SG find-PERF-3.PR.SG “He has found the bear’s lair;” Bi ollomÙcin-i-m diab-då-n I come.to.fish(PERF)-CONN-1.PR.SG boat-DAT-3.POSS.SG am§k§-v. grandfather-1.POSS.SG “I went fishing in my grandfather’s boat” (KOL:38-9). In the Eastern Evenki dialects the first type of “reflection”can be observed in the attributive word-combinations which consist of personal pronouns for the first and the second persons, singular and plural and a noun: bi pÙta-v I bag-1.POSS.SG “my bag;” si gule-s you house-2.POSS.SG “your house” (Kolesnikova, 1966:38). In fact, the term “reflection” was introduced to designate a variety of attributive word-combinations close to those which were referred
gorel-3.p65
389
4/17/02, 12:35 PM
390
part five
to in Turkic studies as “izafet,” an originally Arab term. The term “izafet” is widely used in Turkology to designate attributive wordcombinations, both components of which are expressed by nouns. Specialists in Turkology normally distinguish three types of attribu tive constructions (word-combinations) components of which are linked by means of “izafet.” Each of these types differs not only morphologically but also with respect to the semantic and syntactic relationship between the components of a construction. The attribu tive word-combination constituted according to “izafet I,” is used to denote qualitative characteristics attributed by the dependent component to the main one. Both components of the construction have no morphological formants. The dependent component of the word-combination formed according to “izafet II,” is used to attrubute relative characteristics to the governing component which is followed by the possessive suffix of the third person, for example: tu�rk dil-i “the Turkish language” (Turkish). “Izafet III” is used to denote possessive relationships in the broad sense; the dependent component is followed by the formant of the genitive, and the gov erning component is followed by the possessive suffix of the third person, for example: /-�O� �'��-! “the death of a horse” (Turkmen) (LE, 1990:172). Some turkologists distinguish “izafet IV” which is also used to denote relative-posessive characteristics. In this case only the dependent component is morphologically marked; it is followed by the marker of the genitive. The attributive word-combinations formed by means of the first type of “reflection” in Tungusic, is very close to those formed by means of “izafet II” in Turkic. Turkologists have different opinions regarding the definition of the grammatical means by which the “izafet-constructions” are formed. Some do not give a special defi nition at all (Maizel, 1957). Others define the grammatical means of this syntactic relationship as a kind of government (Kononov, 1956:411-2). Some view “izafet” as a special kind of agreement (Ubryatova, 1950:39, 43). It should be noted that “reflection,” with some variations, is widely used in all Tungusic languages except Manchu. Instead of “izafet II,” common to all Tungusic, in Manchu the third and the fourth types of “izafet” are widely used. The third type of “izafet” is com monly used in participial constructions which contain their own subject expressed by a noun in the form of genitive. Being a head of a construction, the participle is followed by the suffix -ngge which
gorel-3.p65
390
4/17/02, 12:35 PM
391
syntax
can be defined as a nominalizer in classical Manchu, however, it’s origin can be traced to the possessive suffix. These participial con structions have already been partly described in Morphology (see Section 5.6.5). They are analyzed further in Syntax of complex sentences (see Sections 5.3.3 & 5.4.3). The fourth type of “izafet” is used in attributive word-combinations where the genitive realizes its possessive function (see Part IV, Section 2.8). According to contemporary ideas (notions) of general morphol ogy, “izafet” is a morphological means for marking syntactic subor dination in the structure of a head noun within a noun syntagma. The head noun contains a morphological marker which signals that this head noun has a certain dependent element. This strategy of expressing the dependent syntactic status of a noun is opposed to the category of the case, which marks the syntactic subordination in the structure of a dependent noun within a verbal-noun (predicative-argument) syntagma (Plungyan, 2000:184-5). The op position of these two types of encoding syntactic relationships was initially introduced by J. Nichols and designated by the terms “head marking” vs. “dependent-marking” (Nichols, 1986:56-119). Plungyan believes that two types of “izafet” exist in Turkic languages. A sub ordinate (dependent) noun may occur in the form of the nomina tive or in the form of the genitive while a head noun always con tains a morphological marker which coincides with the possessive suffix of the third person and which signals that this head noun has a subordinate noun. The absence of the genitive case marker indi cates that the subordinate noun has no a definite or specific refer ence (Plungyan, 2000:186). When this morphological device is seen in this way, “izafet I” and “izafet IV” cannot be considered as the “head-marking” type of encoding syntactic subordination (depen dence). Agreement and government are also the kinds of subordinating relationships which are expressed morphologically. In the case of agreement, it is a repetition of one, two or several grammemes belonging to the governing word (determinandum) in the structure of the dependent word (determinatum). Agreement is exclusively used to express attributive relationships in Evenki and Even (Lamut) (Kolesnikova, 1966:43). The other Tungus-Manchu languages in cluding classical Manchu do not use the agreement. In the case of government, the governing word requires certain grammemes from the subordinate word. These grammemes do not repeat those be
gorel-3.p65
391
4/17/02, 12:35 PM
392
part five
longing to the governing word. Government is widely used to ex press subordinating relationships in Manchu. Since juxtaposition has no morphological expression, it is usually interpreted as the most simple subordinating relationship. Juxtaposition is widely used in Manchu as well. Coordination is a grammatical expression of both semantically and syntactically symmetrical relationships between components. De spite the close semantic relationship between components, the lat ter are independent grammatically since there is no formal expres sion of subordination of one component to the other. Coordinative relationships in the Altaic languages (including Tungus-Manchu) differ from those in inflexional languages, for instance, in Russian, where there is no such phenomenon as group grammatical expres sion of the parts of the sentence. Formal specificity becomes clearer when coordinating conjunctions appear. These conjunctions link components of the same type although the type itself can be under stood very widely. Morphological parallelism of forms without con junctions can be considered as the formal grammatical means of expressing coordination. In classical Manchu, differentiation of the parts of speech was in the stage of formation. As already shown, it is very difficult to di vide all nominal words into grammatical classes (see Part IV, Sec tion 2). There are also many nouns and verbs that have the same stems. As a result of weak differentiation of the parts of speech, the parts of a sentence have become of greater importance for the or ganization of the Manchu utterance. Word-order is an important grammatical means to distinguish the parts of a sentence. In the Manchu language the general rule of wordorder is to the effect that the subordinate component of a syntactic unit always stands before the governing one. The attribute stands before the noun on which it depends, the object precedes its gov erning verb, and the subject stands before the predicate. Therefore, if the subject and the predicate have the same syntactic structure, their word-order is the only possible way of distinguishing them. The predicate occupies the final position in the sentence. Subordinate predicates precede the principal ones and take suffixes of the nonfinite forms, e.g. converbs or participles followed by the case mark ers. Relationships between subordinate and principal parts of a com plex sentence are expressed by special syntactic devices which are
gorel-3.p65
392
4/17/02, 12:35 PM
393
syntax
used to determine and specify the syntactic division of a sentence and substitute the weakly developed system of the distinctive parts of speech. Besides word-order, the most important devices which serve to express subordinating syntactic relationships between words are the following: government, juxtaposition, correlation, and par allelism. The variety of attributive constructions formed by means of “izafet III” and “izafet IV” is also of great importance in Manchu, however, the structure of these constructions is different to the ones in the other Tungusic languages.
2. Grammatical Means of Expressing Syntactic Relationships in Manchu 1) Word-order is the basic and obligatory means to distinguish the parts of a sentence. The position of a part of a sentence, as in the other Altaic languages, is governed by the general rule according to which a dependent component always precedes a principal one. Word-order is often the only way to distinguish the leading parts of the sentence—a subject and a predicate— when they are rendered by a syntactic group with the same grammatical structure. A direct object is usually placed between leading parts of the sentence, a subject and a predicate. Despite the fact that the direct object has the strongest bond with the predicate, it may occupy one of three possible positions relative to the predicate. a) The direct object may stand directly before the predicate: donji-ci si te manju bithe taci-mbi se-mbi hear-CONV you(SG) now Manchu book learn-IMPF say-IMPF The imperfect form of the verb se- “say” is used to mark the closing of the direct speech. “So I hear you are studying Manchu now, eh?” (MOLL:15; TM); bi hergen be ara-mbi be letter ACC write-IMPF
“I write letters” (PASH:54);
uthai cacari maikan be ca-fi …
immediately tent tent ACC pitch-CONV
“(They) immediately pitched (their) tents …” (NSB:112).
b) It may be separated from the predicate by an adverbial modifier
of manner, an indirect object, or direct speech:
gorel-3.p65
393
4/17/02, 12:35 PM
394
part five
bithe be ure-me håla book/letter ACC be.familiar.with-CONV read(IMP)
bithe håla- “to study” (NL:31);
“Study (the writing system) thoroughly” (PASH:54);
anculan giyahån kuri indahån be saikan hawk a.dog.striped.like.a.tiger dog ACC good i ulebu-fi … belhe GEN feed-CONV prepare(IMP) anculan giyahån “hawk” (NL:17); “Feed hawks and dogs rather well” (NSB:112); min-i gisun be ama eme de getuken I(bi/min-)-GEN = my word ACC father mother DAT clear fondo ula-ra-o i GEN completely pass.to-PART-INT The form -rao is used to denote the imperative here. “Please, pass on my words to (my) father and mother clearly and in full” (NSB:113). c) It may even precede a subject: tere niyalma be bi akda-mbi that man ACC I trust-IMPF
“I trust that man” (PASH:54).
An indirect object normally stands before a predicate:
ere niyalma de bu-he this man DAT give-PART “(He) gave (something) to this man” (PASH2:3; AH); sa-ci uthai sin-de ala-mbi dere know-CONV then you(si/sin- SG)-DAT tell-IMPF MDL.PTL (probably/likely) “If (I) knew, (I) should tell you” (PASH2:81; QW). An adverbial modifier of time normally precedes the subject, but may follow it: sikse si aibide gene-he bi-he yesterday you(SG) where go-PART be-PART “Where have you been yesterday?” (PSH2:91; QW); jai inenggi sergådai fiyanggå ama eme fakca-ra next day PN PN father mother leave-PART doro-i hengkile-fi … rite-GEN kowtow-CONV fiyanggå “youngest,” “least,” “smallest” (NL:89);
gorel-3.p65
394
4/17/02, 12:35 PM
395
syntax
“The next day Sergådai Fiyanggå kowtowed to (his) father and mother following the parting ritual …” (NSB:112); muse cimari abala-me tuci-mbi we (INCL) morning hunt-CONV leave-IMPF “Tomorrow morning we will go hunting” (NSB:112). An adverbial modifier of place may precede a subject, but more often comes after it: sin-i boo aibide te-hebi you(si/sin- SG)-GEN = your house where sit/reside/be-PAST “Where is your house?” (PASH2:143; QW); age si ubade te-hebi-o elder.brother/sir you(SG) at.this.place/here reside-PAST-INT “Do you live here (at this place; in this neighbourhood—MOLL:36), sir?” (PASH2:172; TM); nahan de emu funiyehe sara-ka sakda oven-bed DAT one hair become.white-PART old mama te-hebi grandmother sit-PAST
“A grey-haired old woman was sitting on a nahan (oven-bed used
in Manchuria; chin. kang)” (NSB:119).
An attribute always precedes the noun it modifies, and rearrangment
is impossible:
min-i beye fulgiyan fi i emke
I(bi/min-)-GEN = my self red brush (< chin. bi) GEN one
emken i kimci-me dasa-ha
one GEN check-CONV correct-PART
emke emken i “one by one,” “little by little” (NL:74);
“Checking (manuscripts) one by one, I myself corrected (mistakes)
with a red brush” (PASH2:329; BB);
suru morin de yalu-fi …
white(horse) horse DAT ride-CONV
“(He) mounted a white horse” (NSB:112).
As one can see from the above examples, a subject normally pre
cedes the predicate which occupies the last position in a sentence.
Dependent predicates always precede the principal ones and sub ordinate clauses always precede the principal ones in complex sentenses. Positions of the parts of the sentence, in addition to word-order, may be marked in some cases by particles. Thus, certain particles
gorel-3.p65
395
4/17/02, 12:35 PM
396
part five
are used to mark a subject. Different particles are used with other parts of a sentence, viz. predicates, objects or attributes. In contrast to the other Tungusic languages, the Manchu sentence may include the so-called thematic subject which normally precedes a grammatical subject. This thematic subject is often followed by special particles which are described in detail below (see Section 4.1.1). 2) Government, as a grammatical means, is used to express subor dinating relationships when the governing component of a wordcombination requires a certain grammatical form (or forms) from the dependent component. The grammatical form of the latter is different from that of the governing component. Governing words determine the grammatical form of subordinate ones. In Manchu, different classes of words with semantics of action, e.g. verbs, par ticiples, converbs, and verbal nouns, can function as governing words. When the relationship between governing and subordinate words is obligatory the government can be defined as strong. When the relationship is not obligatory the government can be defined as weak. Ability to govern strongly is closely related to direct and indirect transitivity. Governing words which possess the category of transi tivity, cannot be used without subordinated words in certain cases. Such sentences as “I weile-mbi” (“He builds”), “ I gai-mbi” (“He takes”) can be understood as incomplete. Complete sentences are: i boo be weile-mbi he house ACC build-IMPF “He builds a house;” i bithe be gai-mbi he book ACC take-IMPR “He takes a book.” Words can be characterized by weak government if they do not possess the category of transitivity. Sentences in which they are used are always complete: min-i gucu ji-he I(bi/min-) = my friend came-PART “My friend came.” The position of a subordinate word (or a group of words, or a de pendent clause) as well as case markers are the formal devices to
gorel-3.p65
396
4/17/02, 12:35 PM
397
syntax
express government. Case markers may sometimes be omitted. In Manchu, postpositions may also govern subordinate elements. Here are some examples: bi bithe be håla-mbi I book ACC read-IMPF “I read a book” (“I am reading a book”); bi tere niyalma de hendu-he I that person DAT say-PART “I said to that person” (PASH2:3; AH); bi boo-ci tuci-ke I house-ABL go.out-PART “I went out of the house;” da-ci dube-de isitala beginning-ABL end-DAT up.to “from the beginning to the end” (PASH2:22; AH). Subordinate elements may serve as direct or indirect objects and adverbial modifiers. They may be rendered by a single word, or an expanded syntactic group, or subordinate clause, or direct speech. Here are some examples: cira be taka-ra gojime mujilen be face ACC know-PART although heart ACC sa-rkå know-PART(NEG)
“Although (I) know the face (of a person) (I) do not know (his) heart
(thoughts);”
yasa-i tuwa-ra-ngge tasara-ha ba bi-sire be eye-GEN look-PART-NR be.wrong-PART SBSTR be-PART ACC sa-ci aca-mbi know-CONV meet-IMPF Tv-ci aca- = should, ought, must; sa-ci aca- “(one) should know;” “(One) should know what you see with your eyes is wrong (mislead ing)” (PASH1:8). The last sentence shows that the form of the accusative “be” is of syntactic nature since it serves here to mark the direct object which has the structure of a sentence (clause): yasai tuwarangge tasara ba bisire
gorel-3.p65
397
4/17/02, 12:35 PM
398
part five
be “what (you) see with (your) eyes is misleading.” 3) As mentioned above (see Section 1), “izafet” is a very specific kind of expression of the subordinating relationship in the Altaic languages (it is worth reminding here that the term “Altaic” is un derstood in this book to encompass the typological affinity of the languages). Most of them have the morphological category of pos session which is used to express this kind of subordinating relation ship. The category of possession is expressed by personal-possessive suffixes. The primary function of these suffixes is to denote the possessive semantics which in some cases may be treated as a type of attributive semantics. Because the morphological category of possession is the most fundamental in Altaic, special care must be taken to explain how it manifests itself in word-combinations of different semantic nature. Possessive relations link nouns which denote an owner and an object which is in the owner’s possession, for example, in Evenki: amin avun-i-n “father’s hat” [father hathe/his]. They may link nouns which designate the whole and the part: evenk. hargi dyapka-n “ edge of a forest” [forest edge-it/its], as well as a unit of measure and an object which is measured. Relationships between persons according to their kinship and sub ordination may also be expressed by the possessive suffixes, for example, in Evenki: hute aki n-i-n “a child’s elder brother” [child elder brother-he/his]. Even word-combinations, where the first el ement can be put in correlation with an agent and the second de notes the action which the agent performs, may be organized struc turally by means of the possessive construction. In this case the semantics of a possessive word-combination is very close to that of a predicative construction. As a matter of fact, such word-combinations, formally possessive, are predicative according to their con tent (Avrorin, 1959). These possessive constructions are widely spread in the Turkic, Mongolian, and Tungusic languages. The following are examples of word-combinations where the predicative relation ship between two components is manifested by the form of a pos sessive construction: Evenki: Tadå baka-ra-n untal-va bu there find-PERF-3.SG.PR boots-ACC we
gorel-3.p65
398
4/17/02, 12:35 PM
399
syntax
o-n§-v§-vun.
make-PART-ACC-1.PL.POSS
“There he found (special Tungusic) boots which were made by us;”
Bu mervun sine-ve eme-ne-du-s we ourselves you-ACC come-PART-DAT-2.SG.POSS urun-e-v. rejoice-PERF-1.PL.PF “We rejoiced that you came (We rejoiced at your coming)”1. In classical Manchu there is no the morphological category of pos session. Presumably in the old language this category was expressed by the suffix -ngge, originally of possessive nature. The possessive meaning of the suffix -ngge is seen even in the classical language where it acts as a nominalizer rather than a possessive suffix. Quite often participles may be followed by the suffix -ngge which reveals its possessive nature in the following constructions: ere sargan i ji-dere-ngge this women GEN come-PART-NR “this women’s approach;” ejen i buce-he-ngge lord GEN die-PART-NR “the khan’s death;” min-i ubaliyambu-ha-ngge I(bi/min-)-GEN = my translate-PART-NR “what I translated,” “my translations;” sin-i taci-ha-ngge you(si/sin- SG)-GEN = your learn-PART-NR “your learning.” Thus, the possessive relationships between an agent and an action performed by him/her, may be rendered by the construction “noun/ pronoun GEN Tv-PART-ngge.” Formally possessive, they are predicative in content. When such constructions have no comple mentary components describing the situation in a more detailed way, participles can rightfully be translated as nouns of action. When they are expanded with dependent words these constructions obtain the
1
The bar over letters is used in the examples from Evenki to indicate long vowels.
gorel-3.p65
399
4/17/02, 12:35 PM
400
part five
ability to reflect certain situations of the outside world in more detail and therefore have good reasons to be classified and translated as subordinate clauses. The attributive (possessive) relationships between nouns are nor mally expressed by the genitive in Manchu. The following are ex amples of nominal word-combinations where the genetive is used to denote the attributive (possessive) relationships: moo i ejen “the owner of a tree;” han i sargan “the khan’s wife;” abdaha i amtan “the smell of a leaf;” abkai elden “sky light” (the light from the sky); juse i usabun “affection towards children.” There are constructions of this type where possessive relationships link not two nouns, but a noun and a participle. In this case the first component of the possesive construction may be regarded as an agent and the second as a predi cate. Formally possessive, these relationships, according to their content, are transformed into predicative: ama i håla-ra father GEN call-PART “the father’s call;” han i ji-he khan GEN come-PART “the khan’s coming;” sin-i ara-ha you(si/sin- SG)-GEN do-PART “what you have done,” “your action (act).” 4) Juxtaposition is defined as a kind of grammatical means of ex pressing subordinating relationships when the form of the depen dent component of a word-combination does not depend on a gov erning component and does not change according to its requirements. It is known that jaxtaposition is more typical for analytical languages or for those that have analytical features, for example, Turkic. In Manchu, jaxtaposition is often the only possible device to combine components into word-combinations, for instance, an attribute and a determined member, an adverbial modifier and a predicate. The word-combinations “sain yabumbi” (“one acts well”) and “sain baita” (“good matter”) are distinguished depending on whether a subordi nate component combines with a verb or a noun. In these examples the syntactic position is the only device to express, in the first case, the adverbial modifier of manner, and, in the second case, the at tribute.
gorel-3.p65
400
4/17/02, 12:35 PM
401
syntax
In Manchu, the ability to be combined with governing words by means of juxtaposition, is typical for nouns with the meaning of quality, as well as for demonstrative, interrogative or indefinite pro nouns and participles when they serve as attributes. Serving as adverbial modifiers, adverbs and converbs are also linked to gov erning words by means of juxaposition. A peculiarity of juxtaposi tion regarding grammatical characteristics of dependent words is that some of them, nouns and adverbs, are used in their invariable forms, but others, converbs and participles, which in fact are the special verbal forms, express the subordinating relationship by their own forms. In other words, the formally dominating verb requires spe cial converbal or participial forms from the dependent component. The relationships between components, which are linked in wordcombinations by means of juxtaposition, can vary in respect to close ness. Thus, a tighter link is typical for the relationship between adverbial modifiers of manner expressed by qualitative or quantita tive adverbs, and predicates. In this case adverbs characterize the action in regards to its quality or quantity. Adverbial modifiers of time and place characterize a sentence as a whole, therefore their relationships with predicates are not so tight. In some cases syntactic word-combinations which have attribu tive meaning and components of which are linked by means of jux taposition, become lexical words: siden haha “juvenile,” “minor” (siden “interval,” haha “man”). Sometimes it is difficult to distinguish jux taposition from collateral subordination: ama jui i boo father son GEN house “the house of a father and a son;” ama-i jui i boo father-GEN child GEN house “the house of the father’s son;” sain ehe be yabu-mbi good evil ACC make-IMPR “One makes good and evil.” 5) Correlation is a grammatical means which is used to express the predicative relationship between a subject and a predicate in Manchu sentences. It is mostly limited to nominal predication. From the point of view of general linguistics, the nature of the
gorel-3.p65
401
4/17/02, 12:35 PM
402
part five
predicative relationship, i.e. how a subject and a predicate coordi nate, deserves special discussion. Some scholars consider this rela tionship subordinating, with the subject being a governing member. Others give equal rights to the subject and the predicate. The rela tionship between them is considered neither subordination nor co ordination. In their opinion, the predicative relationship has a spe cial nature when a subject and a predicate formally coordinate with one another. Finally, there is a point of view according to which the relationship between a subject and a predicate is considered subordination, but with the predicate being a governing member. According to this concept, the predicate is the dominant part of the verbal sentence (Tenier, 1988; Kholodovich, 1979:293-8). Some scholars believe that the question about the grammatical nature of the predicative relationship cannot be solved with a help of a for mal criterion. A functional criterion should be involved, and accord ing to this view, a predicate must be considered as a dominant, at least in a verbal sentence. This is a verb-predicate which determines the pattern of such a sentence (Maslov, 1987:178-9). As far as Manchu is concerned, the predicative relationship be tween a subject and a predicate is formally expressed by a particu lar grammatical means which can be called correlation. In the case of correlation, the subject is followed by specific markers which require certain markers from the predicate. As a result, the corre lation between two groups of markers has evolved. Elements of one of them serve to mark the subject (the group of subject), and the others are used to mark the predicate (the group of predicate). This formal expression of syntactic relationships is typical for nominal predication. Thus, the word serengge (originally the imperfect parti ciple followed by the suffix -ngge and derived from the verb se- “to say” in its auxiliary function) requires the copula be in the predicative position. The imperfect and perfect participle, followed by the suffix -ngge, in their positive and negative forms, coordinate with the nega tive copula akå. If a predicate is expressed by the word gånin “thought,” “opinion,” “feeling,” a subject is followed by the word sere, originally the imperfect participle derived from the verb se- in its auxiliary function, etc. Such correlative patterns are discussed in more detail in a section dedicated to nominal predication ( see Section 4.1).
gorel-3.p65
402
4/17/02, 12:35 PM
403
syntax
6) Parallelism is a link between the homogeneous parts, simple and expanded, in a fused sentence as well as between homogeneous collaterally subordinated or coordinated sentences (clauses) in a composite sentence. Parallelism manifests itself by the following features: 1) repeating the parts of a sentence or sentences which are characterized by the same syntactic structure; 2) the parallel parts of a sentence have the same number of words; 3) the parallel parts of a sentence have the same word order; 4) the parallel parts of a sentence should have the same grammatical form, or in the case of different morphological forms, their syntactic structure should have the same meaning. Parallelism can by characterized as total when all four features are present. It may be partial if not all of the fea tures are realized. The presence of the first feature is obligatory for all types of parallelism. Parallelism can be considered a type of coordination. Here are some examples: aisin menggun oci guise de tebu-mbi, gold silver TOP chest DAT put.in-IMPF bele jeku oci tsang de asara-mbi rice grain TOP granary DAT store-IMPF “As for gold and silver, they are put in a chest, but as for rice and grain, they are stored in a granary” (PASH2:3; AH). In this compound sentence, parallelism is realized in its total form: both simple sentences have topic-comment structures. Topics and comments (rhematic parts of sentences), which contain adverbial modifiers of place and predicates, are parallel in both simple sen tences. All parallel members consist of the same number of words: the topic includes two words, and the other parts of the sentence have one word each. Each sentence is characterized by the same word order: the topic stands first, the predicate terminates the sen tence, and the adverbial modifier of place, which belongs to the rhematic part of the sentence, stands between the topic and the predicate. All parallel members have the same grammatical struc ture: topics are followed by the word form oci that is a special topic marker, adverbial modifiers of place are marked by the dative, and both predicates are expressed by the finite imperfect form in -mbi (for data concerning the means of pragmatic structuring in Manchu see Section 3).
gorel-3.p65
403
4/17/02, 12:35 PM
404
part five
baita be deribu-re de, deribun be bodo-mbi, business ACC begin-PART DAT beginning ACC plan-IMPF deribun de kice-fi, dube de beginning DAT concentrate.on-CONV end DAT heolede-rakå be.careless-PART(NEG)
“When starting a business, (he) plans the beginning; concentrated
on the beginning, (he) calculates the end (of the business)”
(PASH2:38; MB).
This compound sentence includes two complex sentences each of
them containing a temporal clause, although expressed differently:
in the first sentence the temporal clause is expressed by the imper
fect participle in the dative, and in the second sentence the tempo
ral clause is rendered by the perfect converb. The principal predi
cate of the first complex sentence is expressed by the imperfect finite
form, but the principal predicate of the second complex sentence is
expressed by the negative form of the imperfect participle.
In addition to the above, parallelism is used as grammatical means
of expressing the comparative degree, for example (see also Part
IV, Section 2.6.1):
ere morin sain, this horse good tere morin sain akå that horse good COP.NEG (there.is.not)
“This horse is better than that” (literally: “This horse is good, that
horse is not good”).
To a certain degree every language possesses a syntactic means such
as parallelism. However, in Manchu parallelism is so widespread due
to translations from Chinese where it is one of the basic grammati
cal and stylistic means.
3. The Means of Pragmatic Structuring in Manchu As it is widely accepted among linguists, the sentence structure can be fully described only as a multilevel hierarchy in terms of differ ent levels of grammar. Various semantic, syntactic, prosodic, and
gorel-3.p65
404
4/17/02, 12:35 PM
405
syntax
discourse characteristics should be taken into consideration in ex ploring the sentence structure. Apart from its formal-syntactic structure, every sentence has an information structure which is studied in discourse pragmatics. The information structure of a sentence, a term introduced by M. Halliday (1967), deals with different grammatical structures of a sentence that appear under different communicative circumstances and reflect “a speaker’s assumptions about the hearer’s state of knowledge and consciousness at the time of an utterance” (Lambrecht, 1994:XIII). According to Lambrecht, the most fundamental categories which engaged in the information structure of a sentence are: presupposi tion and assertion, identifiability and activation, topic and focus (Lambrecht, 1994:6). Every sentence in any specific language has its own information structure which manifests itself through differ ent formal devices. Special morphological markers, certain syntactic-creating and complex constructions, forms of syntactic constitu ents are used to organize the information structure. The position and ordering of these constituents in the sentence, special lexical items; prosodic factors (intonation contour, sentence stress) also play a great role in the formation of the information structure of a sentence (Li & Tompson, 1976; Gundel, 1988; Harlig & Bardovi-Harlig, 1988; Kim, 1988; Lambrecht, 1994; etc.). Any information structure re veals itself within a discourse. However, it is an entity of sentence grammar, not discourse (Lambrecht, 1994:7). It is now commonly accepted to describe the information struc ture of a sentence in terms “theme-rheme,” “topic-comment,” “topic focus,” “topic-secondary topic-focus,” “known vs. un-known/ new,” “discourse-OLD vs. discourse-NEW,” “background-focus”2. Some linguists consider that notions in these pairs roughly corre spond to each other, however, others prefer viewing them as differ ing in certain properties. The binary opposition “theme” vs. ”rheme” and further “topic” vs. “comment” as belonging to the communicative structure of a sentence, was intensively analyzed within the Prague School tradi 2 J. Gundel suggests to distinguish between the pragmatic opposition “topic” vs. “comment” and syntactic opposition “topic” vs. “focus.” According to her, “an expression which refers to the topic (as pragmatic entity—L.G.) and which occupies a syntactic position reserved for topics will be referred to as syntactic topic,” and “that part of the sentence which encodes the comment (as pragmatic entity—L.G.) will be referred to as the focus” (Gundel, 1988:210-11).
gorel-3.p65
405
4/17/02, 12:35 PM
406
part five
tion (Mathesius, 1928 (1964); 1929 (1983); Danes, 1974; Firbas, 1966; 1974; etc.). Theme is an initial, starting point for developing actual information, this is what is more or less known to the addressee of communication. Rheme is the new information which is told about the theme, and it is the main purpose and basic communicative content of the utterance. In the simplest cases the theme may coin cide with the subject of a sentence, and the rheme—with the predi cate. The theme, as a rule, is given by the preceding context and it may be inferred from it. The theme may be omitted without any damage for the understanding of the utterance. The rheme is never omitted and it is considered “as being charged with “communica tive dynamism” (Kim, 1988:148). Normally, the theme stands be fore the rheme in the lineal structure of the sentence. To empha size the rheme, it is placed forward, at the beginning of the sentence. In this case the rheme is characterized by a special intonation con tour, and on the level of communicative content—by a greater degree of expressivity and emotionality. The so-called actual division of a sentence, a term introduced by Mathesius, is normally rendered by intonation and word order in some languages. However, when word order is mostly used as a means of expressing a formal-syntactic division of a sentence, lan guages develop some other means of expressing the actual division of a sentence. Thus, to mark the rheme, emphatic and restrictive particles are often used. Demonstrative pronouns are frequently used to mark the theme, and indefinite ones often mark the rheme. In some languages, special syntactic constructions are also used to denote the theme and the rheme. For example, in English the theme may be rendered by the syntactic construction “as smb./smth. is/are concerned.” The English construction “that is he who ...” is an example of a syntactic construction which is used to indicate the rheme when it is the subject of a sentence. The definite and indefi nite articles can be used to actualize the information. The indefi nite article is often used to indicate the rheme, while the definite article marks the theme. The actual division of a sentence may be expressed by the transformation of voices. For example, in English the transformation of the active voice into passive is often used to express actual information (Maslov, 1987:184-5). The definition (and the notion) of topic has undergone certain changes, from the intuitive characterization of topic as “what the speaker is talking about” (“what a sentence is about”) to a rather
gorel-3.p65
406
4/17/02, 12:35 PM
407
syntax
refined one, proposed by Lambrecht. According to him, the topic “has to do with the pragmatic relation of aboutness between discourse referents and propositions in given discourse context” (Lambrecht, 1994:xiv)3. Nevertheless, the majority of scholars agree that the notion of topic is discourse oriented and context dependent. The topic is a referential entity, and it may be given, known or old (predictable) information. The topic is not determined by the verb, and it is “syn tactically independent of the rest of the sentence.” Nominal phrases (NPs), which are used to render the topic, must be definite or ge neric. The topic tends to occupy the sentence-initial position4. The topic is likely to be a universal entity, however, it may manifest it self in different ways in various languages (Li & Tompson, 1976:X, 463, 465, 484, etc.; Harlig & Bardovi-Harlig, 1988:127; Kim, 1988; Gundel, 1988). From the point of view of the “synchronic” vs. “diachronic” di chotomy, subjects can be analyzed as grammaticalized topics (Li & Tompson, 1976:484). T. Givo( n proposes that subjects are diachronically derived from topics, and W. Lehmann made an at tempt to show that subjects have been a result of their development from topics in Indo-European languages (Givo(n, 1976; Lehmann, 1976). The notion of comment and, further, focus, has been also devel oping, from such definition as “the main predication that is asserted, questioned, etc., relative to the topic” to the definition of topic as “that element in a pragmatically structured proposition whereby the assertion differs from the presupposition and which makes the ut terance of a sentence informative” (Gundel, 1988:210-11; Lambrecht, 1994:xiv, respectively). Nevertheless, the majority of specialists agree that comment (or focus) is context independent, it is not a referential entity, it is re 3 The notion of “aboutness” has been elaborated by other scholars; for example, T. Reinhart studies theme (or topic) in terms of “pragmatic aboutness” (Reinhart, Tanya. Pragmatics and Linguistics: An Analysis of Sentence Topics.—Philosophica 27, 1981:53-93). 4 Topic and focus tend to be arranged in a special linear order, which reflects the quality of information, represented by these entities. This tendency is described by the so-called Information Flow Principle (IFP) (Kim, 1988:149). A. Kim quotes this principle as it is formulated by S. Kuno: “words in a sentence are arranged in such a way that those that represent old, predictable information come first, and those that represent new, unpredictable information last” (Kuno, 1978:54).
gorel-3.p65
407
4/17/02, 12:35 PM
408
part five
lated with new information, and it is indefinite (Harlig & BardoviHarkig, 1988:127; Gundel, 1988:212-3). The notions of topic and focus are based on such fundamental categories as givenness, contrastiveness, definiteness, and referentiality (Chafe, 1976; Givo(n, 1976). Later on, when cross-linguistic studies of languages revealed that the binary opposition of “theme-rheme” and “topic-comment” (or “topic-focus”) is not sufficient to describe the diversity of informa tion structures in existing languages, some other notions, related to those of “topic” and “focus,” have been formulated. These are “rela tive topicality,” “referential predictability,” “thematic continuity,” “first clausal topic,” “secondary clausal topic,” “contrastive focus,” “rhematic focus,” etc. (Kim, 1988; Givo(n, 1976, 1983, 1988; etc.)5. T. Givo(n discovered that topicality has scalar nature, and some scalar dimensions can be measured and quantified in discourse. According to him, there are two major components of topicality, namely, anaphoric topical property of “predictability,” and “topic importance.” These components are referred to as discourse mea surements of topicality, and closely related “with the phonological size of the grammatical device used to code the referent.” It was found that the continuity of topics or referents can be measured in terms of the number of clauses between occurrences, the number of re currences of the referent in the preceding discourse, etc. (Givo(n, 1988:249). The underlying principle for obtaining the code-quantity scale has been formulated by Givo(n as follows: “The less pre dictable /accessible a referent is, the more phonological material will be used to code it” (Givo(n. 1988:249). A few languages of the Altaic family, including contemporary Mongolian and Turkish, as well as Altaic-like Japanese and Korean, have been more or less explored with regard to devices which are used to form their informative structures (Erguvanli, 1984; Kim, 1985a, 1985b; Kuno, 1978). The communicative organization of a sentence is now intensively investigated in Ob-Ugrian studies6. 5 According to Kim, the rhematic focus is such an element of the information structure of a sentence that is, firstly, “the core portion of the rheme,” which may be the Np, analyzed as a lexical replacement of the WH-word, and, secondary, it is an “element, representing new information and highlighted by an intonation order” (Kim, 1988:148). He also suggests a hypothesis that in rigid verb-final languages with SOV word order the focused element occurs immediately before the verb (Kim, 1988). 6 I am indebted to Elena K. Skribnik who got me acquainted with the facts
gorel-3.p65
408
4/17/02, 12:35 PM
409
syntax
However, I am not aware of any Tungus-Manchu studies in this field. The only exception is Pashkov’s “The Manchu Language” which in cludes some data related to the so-called thematic subject (Pashkov, 1963:48-9). This phenomenon, being analyzed from the point of view of a contemporary state of discourse pragmatics, displays clear evi dence that it can be identified with topic (see also Section 4.1.1 in which the subject is discussed). Classical Manchu seems to be a language in which topic-comment (topic-focus) structures play an important role. However, it is not an easy task to discover all patterns used to describe the information structure of a sentence in this language. Firstly, the difficulty lies in the fact that the relevant empirical examination requires to explore “linguistic units larger that a single sentence”7. Although the infor mation structure concerns sentence grammar, it reveals its proper ties within a discourse (Lambrecht, 1994:7). As for Manchu, there is no accurate discovery procedure by which one can distinguish sentences (propositions) from each other within a paragraph, as well as paragraphs within the whole text. On the other hand, there is no procedure by which one can reestablish all properties of a single sentence (proposition) which have been lost in the process of divi sion of fragments (paragraphs, text) into separate sentences (see also Section 5). Secondly, there are no available data from classical Manchu about sentence stress and intonation patterns which are engaged in signaling the pragmatic organization of a sentence. Thirdly, at this stage of investigation, the only possibility is to speak about overtly expressed topics (the comment is always overtly ex pressed) whereas “a pragmatic topic is not always encoded as a syntactic topic” (Gundel, 1988:211). Furthermore, some suggested facts cannot be proved or verified by using the experimental meth odology in classical Manchu. Below I suggest a preliminary approach to the problem. More detailed exploration is yet to be done. of Ob-Ugrian studies in discourse pragmatics. I greatly benefited from reading her paper “The Means of Pragmatic Structuring in Nothern Mansi” (to appear in 2001), which was delivered at the 9th International Congress for Finno-Ugric studies (Tartu, August 2000) and at the Conference “Languages and Peoples in Siberia” (Novosibirsk, October 2000). I had been thinking for some time about devices, which are used to express the communicative structure of a sentence in Manchu, but it was her paper that inspired me to put first results in this field in written form. 7 Quoted in J. Gundel, “Universals of Topic-Comment Structure”, 1976:209.
gorel-3.p65
409
4/17/02, 12:35 PM
410
part five
Manchu is one of those languages in which word order of SOV type, is predominantly engaged in the expression of the formal-syntactic organization of a sentence. Although certain rearrangements are possible, these are not of great importance for the expression of changes in the communicative organization of a sentence. The topic tends to occupy the sentence-initial position in Manchu, which agrees with the statement proposed by Givo(n: “The string-initial position invites the hearer to pay more attention, and thus to store and re trieve the information more efficiently” (Givo(n, 1988:276). Although, as mentioned above, the information structure of a sen tence has not been the subject of a comprehensive investigation in Manchu, it seems to be closely related to the existence of theoreti cally interesting and materially diverse devices. Firstly, in Manchu, a number of special word forms are normally used to mark the topic of a sentence. Following Gundel, I use the term “topic marker” to indicate the status of these word forms8. The following word forms are used the most to mark the topic: oci, seci, seme. Being originally verbal forms (mostly imperfect and condi tional converbs) derived from the verbs o- “to become,” “to be” and se- “to say,” “to call,” “to mean,” they literally render the following meanings: “if being/becoming smb. or smth.,” “if speaking about smb. or smth.” It should be noted that in this particular case the word form seme realizes its conditional meaning (see also Part IV, Section 5.7.1). Synchronically, these words are used to signal the topic of a sen tence. Diachronically, it is obvious that these topic markers devel oped from conditional clauses, “which share with topic the prop erty of being given (presupposed) in relation to the rest of the sentence” (Gundel, 1988:219). Here are some examples: (1) etuhun urse oci ehe be yabu-me fafun be powerful people TOP evil ACC make-CONV law ACC neci-mbi violate-IMPF “As far as the powerful people are concerned (as for powerful people), (they), making evil, violate law” (PASH2:44; MB); 8 In fact, Gundel refers the term “topic marker” (and “focus marker”) to morphemes that are used to mark topic (and focus) (Gundel. 1988:216). I think it is also appropriate to apply this term to word forms, not only to morphemes.
gorel-3.p65
410
4/17/02, 12:35 PM
411
syntax
(2) sektefun oci silun baitala pad.used.for.sitting TOP lynx use(IMP) “As far as a pad is concerned, (you should) use lynx” (PASH1:42); (3) aisin menggun oci guise de tebu-mbi, gold silver TOP chest DAT put.in-IMPF bele jeku oci tsang de asara-mbi rice grain TOP granary DAT store-IMPF “As for gold and silver, (they) are put in a chest, but as for rice and grain, (they) are stored in a granary” (PASH2:3; AH).The following translation is also possible: “As for gold and silver, (people) put (them) in a chest, but as for rice and grain, (people) store (them) in a gra nary.” The dative forms of the words guise “chest” and tsang “granary” are used to indicate rhematic focuses in sentences. These locatives oc cur in the immediately preverbal position. (4) tere seci sin-i deo that/he TOP you(si/sin- SG)-GEN = your younger.brother “As far as he is concerned (as for him), (he) is your younger brother” (ORL:137); (5) uttu o-ho-de teni niyalma seme jalan de so be-PART-DAT = if then person TOP world DAT banji-re de yertecun akå o-mbi live-PART DAT shame there.is.not become-IMPF “If so, then a man, while he lives in the world, will not be ashamed (shame will become non-existent)” or “If so, while a person lives in this world, there are no (reasons) to be ashamed “ (GAB:120). The topic is rendered by the word niyalma “a man,” “a person” followed by the topic marker seme. The grammatical subject is expressed by the word yertecun “shame” semantically related to the word niyalma “a man, “ “a person” (a person—his shame, he is ashamed). What is important to note is that in all these sentences, topics are adjoined to full sentence comments, semantic structures of which imply their own grammatical subjects, co-referential to topics, how ever, these are not overtly expressed. In translations these co-referential constituents are mostly rendered by pronouns.
gorel-3.p65
411
4/17/02, 12:35 PM
412
part five
The word form sere, originally a form of the imperfect participle, may be used to indicate the topic. In this case, it is also used for the following two purposes. Firstly, it is used to denote the modality of desire in combination with the preceding form of the optative mood (Tv-ki se-). Secondly, the word form sere, as a participial form, is used to put the whole verbal expression into a noun position, in this particular case, the position of subject, which is also sentence-initial: (6) doro be getukele-ki se-re gånin doctrine ACC explain-OPT say-PART=TOP thought “A desire to explain the doctrine is (his) intention (thought)” (PASH1:34). The word form serengge (literally “what is told,” “that which is said”), originally a form of the imperfect participle followed by the suffix -ngge (a nominalizer), is also used to mark the topic: (7) gosin se-re-ngge niyalma, jurgan mercy say-PART-NR(TOP) human.being the.right se-re-ngge acabun kai say-PART-NR(TOP) duty COP “As for mercy, (it concerns ) human being, as for the right, (it con cerns) duty” (PASH1:33); (8) menggun jiha se-re-ngge ergen be silver money/coin say-PART-NR(TOP) life ACC uji-re sekiyen nourish-PART source “As far as silver and (copper) coins are concerned (as for silver and coins), they are the source of nourishing a life” (ZAKH:322); (9) e a negative.principle ( yin) positive.principle (yang) serengge, emu adali sukdun kai say-PART-NR(TOP) one same spirit/vital.fluid COP “As for yin and yang, they are equally physical” (GAB:120). The word form sehengge (literally “what was told,” “that which was said”), originally a form of the perfect participle followed by the suffix -ngge (a nominalizer), is also used to mark the topic:
gorel-3.p65
412
4/17/02, 12:35 PM
413
syntax
(10) s�u be taci-mbi se-he-ngge, irgebun literature ACC study-IMPF say-PART-NR(TOP) verse bithe be kice-re be book ACC be.diligent-PART COP “As far the study of literature is concerned, the diligent study of The Book of Songs (chin. shijing) is (the most important thing)” (PASH2:55; JB). In this particular case, the word form sehengge is used to put the whole verb phrase s�u be taci-mbi “to study literature” into the sentence-initial position, i.e. into the position of subject. What is important to note is that when an utterance contains a topic marker, developed from the verb o- “to be,” “to become,” the rhematic part is usually expressed by a verbal phrase (see examples 1, 2, 3). When a topic marker developed from the verb se- “to say,” “to call,” “to mean,” which is partly desemantized, the rhematic part is normally expressed by NPs, accompanied by a copula or without it (see examples 4-10). The primary function of these topical mark ers is to put words of any grammatical classes into the position of subject, and from this point of view they can be analyzed as subject markers which are engaged in the formation of nominal predication (see also Sections 4.1.1 & 4.1.2). The position of subject, which is normally the sentence-initial position in Manchu, is very important from the point of view of the communicative organization of utter ance. This position of a language unit is loaded with the highest com municative rank within a sentence. Secondly, the following syntactic-topic constructions are used to mark the topic in Manchu. 1) The so-called “double subject” constructions. The topic and the subject both occur in such constructions so they can be easily dis tinguished from each other. The topic may be adjoined to the full comment (Li & Tompson, 1976:468; Gundel, 1988:224). The first scholar who made an attempt to describe this phenom enon in Manchu was B.K. Pashkov. He distinguished two subjects, the thematic and grammatical ones. However, the phenomenon was noted much earlier by Iv. Zakharov (Zakharov, 1879:322). According to Pashkov, the thematic subject serves as a theme for the whole sentence which has a grammatical subject as well. Both of them may be rendered by the same word forms, mostly by nomi nal words (including substantives) in the nominative which has zero expression. The obligatory characteristic of the thematic subject is
gorel-3.p65
413
4/17/02, 12:35 PM
414
part five
that it always appears at the very beginning of the sentence. A sub ject can be considered as thematic if a sentence contains a gram matical subject as well. Thus, the thematic subject should be distin guished only in the case when there are two subjects in a sentence, thematic and grammatical. The predicate of a sentence mostly cor relates with the thematic subject through the grammatical one. Otherwise, the whole sentence serves as a predicate correlated with the thematic subject (Pashkov, 1963:48-9). The example is as fol lows: (11) tasha (thematic subject/TOP) juwan juwe (grammatical subject) tiger ten two ilaci-de bi third-DAT COP “A three year old tiger looks like a twelve year old one” (ORL:147). 2) The second type is usually referred to as left dislocational syntactic-topic constructions. In this case the sentence contains a pronomi nal form which is co-referential with the NP used to express the topic (Gundel, 1988:224). In Manchu, the demonstrative pronouns ere “this” or tere “that,” as well the word beye in the meaning of the reflexive pronoun “self,” are normally used in the role of such a pronominal element. In this case, these pronouns realize the anaphorical function (about pronominal use of the word beye “body” see Part IV, Section 4.7): (12) hiyoos�un deocin serengge filial duty.of.a.younger.brother say-PART-NR(TOP) tere (grammatical subject) gosin be yabu-re fulehe dere that mercy ACC act-PART root COP “Respect for parents and younger brothers, this is the foundation
of displaying mercy” (PASH1:33).
In this particular case, the topic is also followed by the topic marker
serengge. (13) niyalma de takåra-bu-re niyalma (thematic subject/ man DAT send-PASS-PART man TOP) beye (grammatical subject) suila-mbi, self be.exhausted-IMPF
gorel-3.p65
414
4/17/02, 12:35 PM
415
syntax
niyalma be takå-ra man ACC send-PART
niyalma (thematic subject/TOP) gånin (grammatical subject)
man mind
jobo-mbi worry-IMPF “A man, who is sent by other people, is exhausted physically, a man who sends other people, (his) mind worries” (PASH1:31). The com ment of the second part of this sentence includes the grammatical subject gånin “mind,” which is semantically related to and denotes only one feature of the topic expressed by the word niyalma “a man,” “a person” (a man—his mind). There are specific constructions in which the topic correlates with the grammatical subject through the demonstrative pronouns ere “this” or tere “that” in the form of the ablative: (14) kooli (thematic subject/TOP) ere-ci sain law this-ABL good ningge (grammatical subject) akå, SBSTR COP.NEG (15) gånin (thematic subject/TOP) ere-ci jiramin thought this-ABL thick ningge (grammatical subject) akå SBSTR COP.NEG “There is nothing better than a law, and there is nothing more profound than a thought” (PASH1:31). 3) This type includes the so-called subject-creating constructions, mostly passive in Manchu. As a result of passive transformations, the direct object is normally promoted to the subject position and, as a result, its communicative rank (topicality) increases: (16) ere-be tuwa-ci mujin bisi-re-ngge baita this-ACC look.at-CONV resolution be-PART-NR thing jiduji mute-bu-mbi finally can-PASS-IMPF “If a man but resolve, the thing (he wants to do) is done” (MOLL:44). Pashkov translates this sentence literary as “Looking at this (example), (then we can conclude that the following proverb is true): when/if
gorel-3.p65
415
4/17/02, 12:35 PM
416
part five
there is a resolution, the thing can be done” (PASH2:186). 4) Syntactic subject-creative constructions may be formed accord ing to the following pattern: S {Tv-PART + -ngge} == P {NP (COP)}9. The suffix -ngge, which has been earlier defined as a nominalizer, plays here a crucial role. It is used to put participles and participial phrases into the position of subject (see also Sections 4.1.1, 4.1.2 & 5.4.3). The suffix -ngge, originally possessive, is closely related to the cat egory of possessivity. This category deals with referentiality / speci ficity, and therefore plays an important role in the organization of the informative structure of a sentence. Participles, in their positive and negative forms, followed by the suffix -ngge, are permitted to occupy nominal positions in the sen tence, mostly the sentence-initial position, which is the position of subject. This is the position of a sentence constituent with the high est communicative rank in the Manchu sentence. Hence the suffix -ngge can be referred to as rank-ordering of topic-marking devices. Occurring alone, these participles are normally translated as verbal nouns: (17) niyamniya-ra-ngge gabta-ra-ngge shoot.from.horseback-PART-NR(TOP) shoot-PART-NR(TOP) gemu sain all good “Mounted shooting and unmounted shooting, they are both good;” (18) wesi-re forgos�o-ro-ngge umesi hådun advance(in rank)-PART transfer-PART-NR(TOP) very quick wesire forgos�oro “promotion” (NL:308); “As far as promotions are concerned, they are very quick” (ZAKH:76). Expanded with their own direct objects (and subjects), these parti ciples play the role of the predicative subject which is topicalized:
9 Constructions of this type are referred to as cleft-constructions by Gundel (Gundel, 1988:224). I analyze them in different way, as polypredicative constructions with predicative subjects (see Sections 4.1.2. & 5.4.3).
gorel-3.p65
416
4/17/02, 12:35 PM
417
syntax
(19) terni urgun jili be that(GEN) happiness anger ACC tuembu-hekå-ngge uttu be.revealed-PART(NEG)-NR(TOP) like.this terni < tere/ter + ni = genitive; “The fact that neither his happiness nor anger was revealed, (was really true) like this” (ORL: 176); (20) baibi jete-rakå-ngge (thematic subject) ere-ci for.nothing eat-PART(NEG)-NR-TOP this-ABL amba ningge (gramm. subject) bi-o big SBSTR COP-INT “Among those who earn their bread, are there any bigger than he?! (literally: “Among those who do not eat for nothing, are there any bigger than he?!”) (PASH1:37); (21) min’i ji-he-nge (ma. ji-he-ngge) umusi sain baita I(bi/min-)-GEN come-PART-NR(TOP) very good matter “The fact that I came, is a very good thing;” (22) suen’i (ma. suweni) tere uhuri (ma. uheri) weile-me you(suwe/suwen- PL)-GEN that taken.as.a whole work-CONV ara-ha-nge (ma. ara-ha-ngge) gemu sain o-kini make- PART-NR(TOP) all/everything good be-IMP “Everything that you have made (taken as a whole), may it be good” (SK). Predicative heads of these constructions, expressed by participles in -ngge-form, can be translated by the following expressions: “he who … ,” “she who … ,” “those who … ,” “the fact that … ,” “that which … ”. The suffix -ngge should be specified in such construc tions not only as a nominalizer, but also as a rank-ordering of topicforming markers. The rhematic part of these constructions is nor mally expressed by NPs, together or without a copula (semantic classification of predicates, used to render nominal predication, is represented in Sections 4.1.2 & 5.4.3). Participles in -ngge-form may be followed by the accusative marker. In the Sibe dialect, there are many such forms (Lebedeva & Gorelova, 1994:70-1). When occurring alone, such participles can be analyzed as predicative objects. When expanded with dependent words, they
gorel-3.p65
417
4/17/02, 12:35 PM
418
part five
can be viewed as heads of predicative constructions, which also function as predicative objects. What is important to note is that participles, followed only by the accusative marker, without the suffix -ngge, normally function as predicative objects too. The following question arises naturally: if there is any difference in the meanings of the two forms, Tv-PART + ACC and Tv-PART-ngge + ACC. Let us consider several examples: (23) boo-de dasa-bu-ha-ngge be house-DAT correct-CAUS-PART-NR(TOP) ACC gurun de tucibu-mbi country/ruling.house DAT present.to-IMPF “(Someone) presents to the Court a submission how to improve regulations in the ruling house” (PASH2:29; OB); (24) age si men-de tacibu-re-ngge sir you(SG) we(be/men- EXCL)-DAT teach-PART-NR(TOP) be we se-mbi ACC who mean-IMPF “Sir, for whom do you take (that person) who teaches us?” (PASH2:182). Mo�llendorff gives another variant of this sentence: age si men-de tacibu-re niyalma be sir you(SG) we(be/men- EXCL)-DAT teach-PART man ACC we se-mbi who say/call-IMPF “(I see) you think that it is a regular professor that teaches us” (MOLL:41). As one can see, the same meaning is expressed by two different forms. In the first variant, the participle occurs in the -ngge-form before the form of the accusative, and in the second version, the participle is only followed by the accusative marker. (25) geren haf-sa (ma. hafasa) cooha nirgen (ma. irgen) ere sargan’i many official(hafan)-PL soldier people this woman jide-re-nge (ma. jide-re-ngge) come-PART-NR(TOP) be goroki-ci sabu-fi hendu-me… ACC distant/distant.place-ABL see-CONV say-CONV
gorel-3.p65
418
4/17/02, 12:35 PM
419
syntax
“Many officials and soldiers saw from the distance that this woman was approaching and said …” (SK:71). In all these sentences, participial forms in -ngge encode direct objects, which are thematically more important than subjects, i.e. have the highest communicative rank in the sentence. They are also referentially more predictable, which becomes obvious from the syntactic context. However, we do not observe the promotion of directs objects, predicative in their nature, into the position of sub ject. 5) Syntactic subject-creative construction may be formed according to the following pattern: S{Tv-PART + ba} == P {NP (COP)}. The syntactic element ba, which was earlier defined as substantivizer, is used to place a participial phrase into the position of subject. (26) manju bithe-de untuhun hergen baitala-ra ba umesi Manchu book-DAT empty letter use-PART SBSTR very labdu many “As far as auxiliary particles are concerned, in Manchu books there are many of (them)” (PASH2:51; JB). Placed in the position of sub ject, the whole participial phrase acquires at the same time the sta tus of the highest communicative rank within the sentence.
4. Syntax of the Simple Sentence The simple sentence is a binary syntactic unit where two principal members, a subject and a predicate, are put into correlation. There are two types of predicative relationships between them, viz. nomi nal and verbal predication. In the first case, a predicate is normally expressed by a nominal word accompanied by a copula or appear ing without it. A predicate can also be expressed by a copula with out a nominal word. In the second case, a predicate is expressed by a verb, or a verb combination, or an analytical verbal form. 4.1. Grammatical Structure of a Nominal Sentence The most widespread pattern of a simple sentence where the prin cipal parts are linked by nominal predication, is the following one: S (N) = P{N (+ COP)}/{COP}.
gorel-3.p65
419
4/17/02, 12:35 PM
420
part five
In a general case, a nominal predicate includes two structural po sitions, one of which is filled by a nominal word and the other by a copula, for example: muse niyalma bi we(INCL) people COP “We are people” (PASH:50). Being a structural component of the nominal predicate, a copula is combined with a nominal word (or a functional analogue of a nomi nal word) to produce the actual attribute of the subject, i.e. the attribute ascribed to the subject in the utterance. In Manchu, the most widely used copula is bi. Originally the imperfect finite form of the verb o- “to be,” “to become” in its auxiliary use, the word form ombi may also occur as a structural component of a nominal predicate, i.e. a copula. According to B.K. Pashkov, the copula bi represents permanent attributes intrinsic to the subject, while the copula ombi represents temporary ones (Pashkov, 1963:50): senggi fulgiyan bi blood red COP “Blood is red;” mergen ombi wise COP “(He) is wise;” min-de ahån ombi I(bi/min-)-DAT elder.brother COP “(He) is my elder brother” (PASH:50). The copula is frequently omitted, as in: gisun tome ijishån, hergen aname tomorhon word/speech all correct letter one.by.one clear “Every sentence is correct, every letter is clear” (PASH2:161; TM.). Mo�llendorff translated this as: “Every sentence runs as it should; every letter is clear” (MOLL:28); muse ahån we(INCL) brother “We (are) brothers” (PASH1:44); suwen-i gisun umesi giyan you(suwe/suwen- PL)-GEN = your word very true
gorel-3.p65
420
4/17/02, 12:35 PM
421
syntax “Your words are very true” (NSB:115).
When it is required to refer the content of a sentence to the gram matical sphere of the past or the future, the temporal or, to be more precise, aspectual-temporal forms of the verbs bi- and o- are used: taidzu wang o-ho Taidzu monarch be-PART(PERF) “Taidzu was a monarch” (PASH1:44); nis�an saman ambula bayan place-name shaman greatly/enormously rich o-ho become-PART(PERF)
“Nishan shamaness became enormously rich” (NSB:140);
wesihun be sira-ha-ngge ilan wang bi-hebi supreme.power ACC inherit-PART-NR three king be-PAST “Those who inherited supreme power were three kings (princes)” (ORL:176); si emu sain niyalma bi-he-ni you(SG) one good person be-PART(PERF)-EXCLM “You are a really good person!” (NSB:127). Grammatical forms of the verb bi- and o- can be followed by the copula bi: yilan (ma. ilan) sargan jui bi-he bi three female child be-PART(PERF) COP “There were three daughters” (SK). The first component of a nominal predicate can be rendered by any nominal word, alone or extended with complementary words, for example: mudan gai-re-ngge sain bi-me tomorhon pronunciation take-PART-NR good be-CONV = and clear “Your pronunciation has become good and clear” (PASH2:153; TM). Mo�llendorff translated this as “Your pronunciation is good [and you speak quite intelligibly]” (MOLL:21). si aika gurun gåwa-o you if/whether country other-INT
gurun gåwa “a foreigner;”
“You are not a foreigner, are you?” (PASH2:166; TM). Mo�llendorff
gorel-3.p65
421
4/17/02, 12:35 PM
422
part five
translated this as “You are one of us, are you not?” (MOLL:19); umesi baktarakå bayan very extremely rich/rich.man “(He was) extremely rich” (NSB:111). A number of other copulae may function as substitutes of the copu lae bi and ombi, such as kai, be, inu, dere, dabala (for details see Part IV, Section 9). Possessing certain modal meanings, all of them take part in the formation of nominal predication: si nikan bithe bahana-ra niyalma kai you Chinese book comprehend-PART man COP “You are a Chinese scholar” (literally: “You are a man who has studied Chinese books”) (PASH2:159; MOLL:26; TM); ere da sekiyen dabala this source beginning COP “(Certainly) this is the source of all beginnings” (ORL:220); ere gemu hesebun kai this all fate COP
“All this is fate” (NSB:135);
min-i beye weihun gurun i nis�an
I(bi/min-)-GEN = my self alive country GEN place-name
saman inu
shaman COP
“I myself am Nishan shamaness from the country of living (people)”
(NSB:126).
In Manchu there are several negative copulae, such as akå, waka,
unde that also may form predicativity of nominal sentences negating
nominal predicates (for details see Part IV, Section 9). Here are some
examples:
fuhali te-de bakcin waka actually that-DAT match COP(NEG) (is not) “In fact, I am no match for him” or “In fact, I cannot be compared with him” (PASH2:155; TM). Mo�llendorff translated this as: “I am as far from being his match” (MOLL:22). bi saman waka I shaman COP(NEG) (is not)
“I am not a shamaness” (NSB:119);
gorel-3.p65
422
4/17/02, 12:35 PM
423
syntax
ehe fayangga waka bad soul COP(NEG) (is not)
“I have no bad intentions” (literally: “there is no bad soul”)
(NSB:133);
enteke niyalma be sa-ra unde this.sort.of man ACC know-PART COP.NEG (not yet) “(I) have not known this man yet” (literally: “ There has not been yet the knowledge of this man”) (PASH:51). A negative copula can be followed by any grammatical forms of the verbs bi- or o-: age sin-i ere uthai waka elder.brother/sir you(si/sin- SG)-GEN = your this then NEG o-hobi be-PAST “Sir, your (actions) are really wrong” (PASH2:92; QW). From the semantic point of view, sentences with nominal predica tion are represented by two different types. Firstly, these are sen tences where the predicate denotes a permanent attribute of an object (or an event of the outside world), in particular the presence or absence of this object (or event), for example: senggi fulgiyan bi blood red COP
“Blood is red;”
tere anggala, hono emu aldungga ba-bi
that not.only still/yet one strange thing-COP
tere anggala “moreover;”
“Moreover, there is another strange thing” (PASH2:166; TM).
Mo� llendorff translated this as “Then there is another odd thing”
(MOLL:32);
majige nikan mudan akå little Chinese accent COP.NEG (there.is.not)
“There is not even a little Chinese accent” (PASH2:154; TM).
Mo� llendorff translated this as: “He speaks without a particle of
Chinese accent” (MOLL:22);
majige cilcin akå little fault COP.NEG (there.is.not)
“There is not even a little fault” (PASH2:161; TM). Mo� llendorff
translates this as: “I have not a fault to find” (MOLL:28);
gorel-3.p65
423
4/17/02, 12:35 PM
424
part five
urgun be cira de serebu-rekå-ngge akå happiness ACC face DAT reveal-PART-NR COP.NEG (there.is.not) “There do not exist those who do not show happiness on their faces” (ORL:212). On the other hand, these are sentences where an object (or a given event) is characterized through correlation with the categories of human thought. This semantic type includes two groups of sentences: constructions of classification, where the given object is put into correlation with a certain class of objects, and constructions of iden tification, where the objects are identified with each other. From the logical point of view identification is a particular case of classifica tion. Here are some examples: muse niyalma bi we(INCL) people COP “We are people;” ere niyalma sain akå this man good COP.NEG (there.is.not) “This man is not good” (ORL:212); ferguwecuke enduri saman kai wonderful deity shaman COP
“(You) are really a wonderful shamaness” (NSB:121);
tere jide-re-nge (ma. jide-re-ngge) men’i (ma. meni)
that come-PART-NR we(be/men- EXCL)-GEN = our
mama kai
old.woman COP
“The one who is coming is our old woman” (SK).
In sentences with nominal predicativity, the forms of subjects are
put into correlation with the forms of predicates. Before discussing
these forms, special attention should be paid to the category of subject
in Manchu in general.
4.1.1. The Category of Subject The identification of a constituent of a sentence with the subject is one of the most difficult questions of the theory of syntax. As shown in many recent studies, there is no universal definition of subject, and “subjects can vary in their properties even within a specific language” (Li, 1976:ix). Subject is basically seen as a relational notion.
gorel-3.p65
424
4/17/02, 12:35 PM
425
syntax
This notion conveys the grammatical function performed by one of the principal constituents of the “subject-predicate” structures. The majority of the properties of subjects, as listed by E. Keenan, are syntactic in nature (Keenan, 1976). In Manchu, the category of subject is characterized by several dis tinctive features. First of all, as compared with other Tungusic lan guages, such characteristics of subject as number and person are not reflected by the verbal form. Secondly, in sentences subjects are often omitted, especially in their co-referential use, when referents are most predictable. Since semantic and syntactic sentence structures imply subjects, as the important constituent of the sentence, subjects can be easily reconstructed from the syntactic context. However, this re construction requires operating with language units larger than a single proposition. Thirdly, certain Manchu sentences may include a subject and a topic, and in some cases there is no precise discov ery procedure using which one can easily distinguish between them. According to Pashkov, one can speak about the topic (thematic subject) only in case when both of them present in a sentence (Pashkov, 1963:48). In my view, subject-creative constructions can be referred to as the so-called rank-ordering of topic-marking de vices because they are used to increase the communicative rank of language entities placing them into the position of subject. In Manchu these constructions are simultaneously subject-creative and rankordering. In simple verbal sentences the overtly expressed grammatical sub ject may be rendered by the nominative form of a nominal word, a nominal phrase or a pronoun, as in: sin-i boo aibide te-he-bi
you(si/sin- SG) house where sit/live/be-PART-COP
“Where is you house?” (PASH2:143; QW);
gåwa ju-se sabu-fi gemu golo-fi sucu-me
other child( jui)-PL see-CONV all be.scared storm-CONV
boo-de dosi-fi…
house-DAT enter-CONV
“Other kids saw (that) and became scared, all rushed to (their)
home…” (NSB:128);
geren aba i urse morin be dabki-me…
all battue GEN people horse ACC whip.on(a horse)
“All hunters whiped on horses…” (NSB:112);
gorel-3.p65
425
4/17/02, 12:35 PM
426
part five
beile age jugån de nime-me beje dube-he ruler master road DAT be.ill-CONV self die-PART “The master got sick on the way and died” (NSB:114); bi juwan aniya funceme nikan bithe taci-ha I ten year over Chinese book study-PART
funce- “to be in excess;” funceme “over,” “in excess” (NL:95);
“I have been studying Chinese for over ten years” (MOLL: 16);
si uba-de te-he-bi-o you(SG) here-DAT live-PART-COP-INT “Do you live here?” (PASH2:172; TM); sin-i beye gene-ki se-he dabala, we you(si/sin- SG) self go-OPT say(AUX)-PART MDL.PTL who sim-be gene se-he you(si/sin-)-ACC go(IMP) say-PART “You yourself wanted to go, who forced you to go (literally: who told you: go!)?” (PASH2:101; QW). In sentences that are organized by means of nominal predication, subjects may also be expressed by nominal words, nominal phrases or pronouns (see also Section 4.1): bi gulu suwayan ningge I pure yellow SBSTR
gulu suwayan = the pure yellow banner, which is one of the eight
Manchu banners, military organization of the Manchus (NL:114);
“I am of the pure yellow banner” or “I belong to the pure yellow
banner” (PASH2:134; QW);
min-i age ne lamun funggala
I(bi/min-)-GEN sir now blue feather
the word age is used here in the meaning of ahån “elder brother;”
lamun funggala “Junior Bodyguard (wearing the Blue Feather)”
(NL:185);
“My elder brother is now the Blue Feather Junior Bodyguard” (My
elder brother is now one who wears Blue Feather on his hat)”
(PASH2:141; QW);
ajige juse cun cun i mergen haha o-mbi small children gradually wise man become-IMPF chun chun i “gradually,” “by degrees” (NL:48); “Children gradually become wise men” (ORL:147);
gorel-3.p65
426
4/17/02, 12:35 PM
427
syntax
A nominal word may be followed by one of the markers oci, seci, seme, sere, serengge, sehengge which are also used to indicate the topic (theme). In such cases the grammatical subject may coin cide with the theme (see examples 4-10 in Section 3). There is one more example: amban serengge hese be ali-fi wen be high.official THM(TOP) edict ACC receive-CONV education ACC selgiye-re-ngge disseminate-PART-NR “A high official, having received an edict, disseminates education” or “A high official is one who, having received an edict, dissemi nates education” (PASH1:33). To create grammatical subjects, the suffix -ngge is widely used in Manchu sentences. Following participles, standing alone or ex panded by complementary words, this suffix is used to put them into the subject position (for details see Part IV, Sections 2.4.1. & 5.6.5 and Part V, Sections 4.1.2 & 5.4.3). These participial constructions are predicative in nature: buce-re banji-ra-ngge emu erin i anan-de bi die-PART live-PART-NR(TOP) one time GEN sequence-DAT COP “Life and death span a short period of time” (AD:66); baldu bayan i jui sergådai fiyanggå be PN rich.man GEN child PN PN ACC baha-fi gaji-ha-ngge be.able-CONV bring-PART-NR(TOP) ejehen muten ajigen akå (?) skill small COP.NEG (there.is.not)
“The fact that you could bring the Baldu Bayan’s son Sergådai
Fiyanggå, demonstrates remarkable skills” (NSB:137).
As mentioned above, occurring in the sentence-initial position,
participles and participial phrases in -ngge form, acquire the highest
communicative rank (topicality) in a sentence (see examples 17-22
in Section 3).
Both a thematic and a grammatical subject may be expressed by participles followed by the suffix -ngge: sain ehe be yabu-ra-ngge (topic), dorgi-ci good evil ACC make-PART-NR(TOP) inside-ABL
gorel-3.p65
427
4/17/02, 12:35 PM
428
part five
tuci-rakå-ngge (gramm. subject.)
appear-PART(NEG)-NR
akå, mujilen ci deribu-rakå-ngge (grammatical subject)
there.is.not heart ABL begin-PART(NEG)-NR
aku COP.NEG (there.is.not) “In doing good or evil, there is nothing that could not appear from the inner state (of a person), and that could not originate from the heart” (GAB:81). Strictly speaking, most constructions, which contain participles in the form in -ngge, overstep the limits of a simple sentence. Many forms, especially those containing subordinate components, may function as heads of predicative constructions. These constructions themselves semantically and structurally refer to a simple sentence which ren ders a situation or an event of the outside world and functions within a communicative unit of a higher rank. Playing the role of a predicative subject, many participial constructions that contain forms in -ngge should be taken into consideration in the Syntax of polypredicative constructions (see Section 5.4.3). To create grammatical subjects, the marker ningge is also used. This marker is defined as a substantivizer since it is such an element by which nominal words shift their category status to substantives (see Part IV, Section 2.4.1 and Part V, Section 4.1.2). Participles followed by the substantivizer ba, may also serve as grammatical subjects (see example 17 in Section 3). The role of gram matical subject, expressed by participles followed by the substantivizer ba, is also closely related to the expression of the rank-ordering devices. When occurring in the position of subject, a participle, followed by substantivizer ba, acquires the highest communicative rank (topicality). 4.1.2. Correlation Between a Subject and a Predicate in the Sphere of Nominal Predication The question about the character of the predicative relationship between a subject and a predicate has already been discussed in Altaic studies, but it has not be completely solved untill recently (PDP, 1984:108). As for Manchu, this question also deserves special discussion, es pecially in the sphere of nominal predication. As shown above, a
gorel-3.p65
428
4/17/02, 12:35 PM
429
syntax
subject (or a syntactic group of subject) is characterized by special forms which correlate with forms of the nominal predicate (or a syntactic group of predicate). Thus, there are many reasons to state that the predicative relationship between a subject and a predicate is expressed by special correlative forms, and the relationship itself may be called correlation. This term was initially suggested by B.K. Pashkov in his manuscript devoted to studying the simple sentence in Manchu and was then used in further development of Manchu studies (Gorelova, 1988:80 ). I have found a set of sentence patterns in which the predicative relationship can be defined as correlation. 1) The following patterns represent sentences where the predicate denotes a permanent attribute of an event of an object, in particu lar their presence or absence. The subject in such sentences is ex pressed by participles, both imperfect and perfect, in their positive and negative forms, followed by the suffix -ngge: a) S {Tv-ra-ngge/-ha-ngge} === P {N} b) S {Tv-ra-ngge/-ha-ngge} === P{COP} c) S{Tv-ra-ngge/-ha-ngge} === P{N COP (bi, ombi, oho, kai, ...)} (In this pattern a predicate may be represented by a nominal word, or by the copula bi (or some other copulae), or by both of these two structural elements.). d) S{Tv-ra-ngge/-ha-ngge; Tv-rakå-ngge/-hakå-ngge}=== P{akå/waka}. From the synchronic point of view, the suffix -ngge, attached to participles, takes part in the formation of the so-called nominalizations (for details see Part IV, Sections 2.4.1 & 5.6.5). Nominalizations may be predicative or non-predicative. In some languages these two types of nominalizations are formally opposed to each other. For instance, in Russian, non-predicative nominalizations are mostly conveyed by abstract nouns (substantives). However, in the Altaic languages, including Manchu, which has a relatively small number of abstract nouns, the opposition between the two types of nominalizations is not so obvious. In order to determine whether or not a nominalization is predicative, one has to take into account a number of characteristics of the syntactic construction which in cludes a participle in the form in -ngge. Predicative nominalizations have semantics and a formal structure which refer to those of a simple
gorel-3.p65
429
4/17/02, 12:35 PM
430
part five
sentence. When a nominalization is represented by a single parti ciple, it can be considered as non-predicative with a greater prob ability than a nominalization which is represented by a participle expanded with complementary words, especially direct objects. In Manchu there are nominalizations which have a formal struc ture of a sentence, but do not correspond semantically to a situa tion or an event of the outside world. Their semantics may be un derstood as an object in the broad sense, a material object (“thing”) or non-material object (“act,” “matter”). Such nominalizations may be called predicative substantives. The patterns of the first type are normally represented by predicative and non-predicative nominalizations as well as predicative substantives. Here are some examples: aca-ha-ngga tongga meet-PART-NR rare/few “(Our) meetings were infrequent” (PASH2:240; GH); tere durun i fiyana-ra-ngge gemu holo kai that form/shape GEN feign-PART-NR all false COP “(Demonstration of) pretended (friendship) is a lie” (PASH2:87; QW); damu fucihiya-ra-ngge teni majige weihuhen o-ho only cough-PART-NR then a.little light become-PART “As for cough, and even it became a little bit slighter” (PASH2:244; GH); wehe ai jibsi-me iktam-bu-ha-ngge inu sain stone what lie.in.layers collect-PASS-PART-NR really good “(Your) rockery (rock-garden) is really good” (literally: “Collected and laid in layers stones are really good” (PASH2:192; TM); buceli be tunggala-ha se-me golo-fi ghost ACC encounter-PART say(AUX)-CONV be.startled-CONV ergen jocibu-he-ngge gemu bi life lose-PART-NR all COP “There were those who having encountered the ghost, were startled to death” (PANG:94); niyalma untuhuri taci-ra be kice-re-ngge waka a person in.vain study-PART ACC study-PART-NR is.not “It is not true that a person’s efforts to study may be in vain” (“The claim that a person makes efforts to study in vain is not true”) (ORL:214);
gorel-3.p65
430
4/17/02, 12:35 PM
431
syntax
abka-i fejergi irgen tumen jaka be heaven-GEN under people a.myriad thing ACC haira-rakå-ngge akå be.compassionate-PART(NEG)-NR COP.NEG (there is not)
abkai fejergi “all under heaven,” “the universe,” “the world;”
“The claim that people are not compassionate with a myriad of
animals in the world, is wrong” (PASH2:315; MFB);
n’alma (ma. niyalma) ere i adali muda-me
people this/he GEN like return-CONV
ji-he-nge (ma. ji-he-ngge) akå
come-PART-NR COP.NEG (there.is.not)
“None such as he were among those who returned” (SK).
2) The following pattern represents sentences where a given event
(or an object) is put into correlation with a certain class of objects
(constructions of classification) or those where the events (or objects)
are identified with each other.
S{Tv-ra-ngge/-ha-ngge} === P{N bi/kai/inu/dere/ ...}
The sentences of the second semantic type contain predicative sub
stantives. These have the meaning of an object (material or non
material) or a person depending on the semantic class of a noun by
which the main predicate is expressed. Here are some examples:
age-i jombu-re-ngge aisin go i gese elder.brother/sir-GEN suggest-PART-NR gold jasper GEN like gisun kai word COP “What the elder brother (or sir) said are words like gold and jas per” (“The words of the elder brother (sir) are like gold and jasper”) (PANG:94); min-i eigen gai-ha-nge (ma. gai-ha-ngge) emu I(bi/min-)-GEN) husband take-PART-NR one gurgu gaska (ma. gasha) inu wild.animal bird COP “One whom I married was actually a wild animal, in fact, a bird” (literally: “One whom I took as a husband was actually a wild animal, in fact, a bird ” (SK). 3) The third type of sentences are those in which words of different morphological classes assume the position of subject by taking on specific markers. These markers correlate with the form of the predi
gorel-3.p65
431
4/17/02, 12:35 PM
432
part five
cate. The patterns according to which these sentences are formed are as follows: a) S {N serengge} === P {N}; b) S {N serengge/sehengge} === P {N bi/kai/be/dabala}; c) S {N serengge} === P {uthai N bi/be}. The word forms serengge and sehengge are the forms of the imperfect and perfect participles respectively, followed by the suffix (the nominalizer) -ngge, and derived from the verb se- “to say.” Be ing partly desemantized, these forms are used as markers of the subject, and correlate with the nominal predicate which can be expressed by a nominal word, or one of the copulae bi/be/kai/dabala, or a nominal word plus a copula: tere age serengge muse-i fe adaki kai that sir TOP we(INCL)-GEN old neighbour COP se-re-ngge say-PART-NR = that which is said/ what is said =TOP “That gentlman is our old neighbour” (MOLL:43; TM); julge-i ambaki serengge hanja bi-he ancient.times-GEN high.official TOP honest be-PART “As for high officials of ancient times, they were honest” (ZAKH:127); bi serengge teni taci-ha saman I TOP just learn-PART shaman “As for me, I am a shamaness who learned (these skills) very re cently” (NSB:121). As mentioned, the word forms serengge and sehengge are used to put words of any grammatical classes in the position of the subject. Even verbs followed by these forms may serve as subjects: gabta-mbi serengge muse-i manju-sa-i shoot.an.arrow TOP we(INCL)-GEN Manchu-PL-GEN oyonggo baita important thing “As for dismounted archery, it is the most important thing for the Manchu people” (PASH2:202; QW); karula-ki serengge ai gisun repay-OPT TOP what word/speech
gorel-3.p65
432
4/17/02, 12:35 PM
433
syntax
“As for (your) desire to repay (my kindness), don’t mention it (liter ally: what is the talk about!)!” (PASH2:151; QW). Mo�llendorff trans lated this as “Talk of handsome return, indeed!” (MOLL:20); age bithe håla-me gene-ki sehengge sain baita elder.brother/sir book read-CONV go-OPT TOP good thing dabala COP
“Sir, your desire to study is a good idea indeed” (PASH2:184; QW).
Mo�l lendorff translated this as “Your desire to study Manchu is a thing commendable in itself” (MOLL: 43). In the two sentences above, the subject is expressed by the analyti cal construction Tv-ki se- which denotes the modality of desire. This construction is placed in the position of the subject by means of the nominalizer -ngge and therefore can function as a noun analogue in a sentence, in this particular case—as a subject. banin serengge uthai giyan be nature TOP then law COP “Nature is a law” (PASH2:43; MB); weihun gurun i nis�an saman serengge uthai bi alive country GEN place-name shaman TOP then COP “I am Nishan shaman from the country of the living” (NSB:133). These patterns may include the partly desemantized adverb uthai “then,” “at once,” “immediately” which indicates the predicative group. In some variants of these patterns, the word uthai may be omitted. 4) The word uthai is also often used in sentences formed accord ing to the pattern S{N ningge} === P{(uthai) N}, where the subject is expressed by a nominal word with semantics of quality and sub stantivized by the sunstantivizer ningge, as in: ten cokohon ningge uthai alin highest.point vertical SBSTR then mountain “The highest things are mountains” (PASH1:36). 5) The substantivizer ningge is also used in the pattern S{N ningge} === P{akå}, where the subject is expressed by a nominal word with semantics of quality, for example: ere erin ci oyonggo ningge akå this time ABL important SBSTR COP.NEG (there.is.not)
gorel-3.p65
433
4/17/02, 12:35 PM
434
part five
“There is no more important time than this (one)” (PASH1:37). 6) The word form sere, originally the imperfect participle from the verb se-, is used to place words of different grammatical class in the position of the subject. It correlates with the word gånin “thought” serving as the predicate. Such sentences are formed according to the pattern S{Tv-ki sere} === P{gånin}: doro be getukele-ki sere gånin doctrine ACC explain-OPT TOP thought “(He) wants to explain the doctrine” (literally: The desire to explain the doctrine is (his) thought” (PASH1:34). The grammatical explanation is as follows: the analytical construc tion Tv-ki se- denotes the modality of desire; in order to be placed in the position of the subject, i.e. to function as a subject, the aux iliary verb se- takes the form of the imperfect participle (-re). This is the form of the participle which allows the syntactic construction to perform the role of the subject. 7) In Manchu there are sentences which contain two nominalizations. The first of them takes the position of the subject, and the other occupies the position of the predicate. In such constructions, the pattern of identification is transformed into the pattern where the subject and the attribute ascribed to it are equivalent semantically and structurally. The pattern according to which such sentences are formed is the following: S{Tv-ra-ngge/-ha-ngge} === P{Tv-ra-ngge/-ha-ngge (kai)}. Here is an example: ama jui be tanta-ra-ngge, jui be gosi-ra-ngge kai father son ACC beat-PART-NR son ACC love-PART-NR COP “The father who beats a son is actually he who loves a son” (PASH:57). There is another variant of this pattern: S {N serengge} === P {Tvra-ngge (kai)}. Here are some examples: amba-sa saisa-i hafan serengge jurgan be high.official-PL gentleman-GEN officer TOP duty ACC yabu-re-ngge kai perform-PART-NR COP
gorel-3.p65
434
4/17/02, 12:35 PM
435
syntax
ambasa saisa “a worthy, wise man, a true gentleman” (NL:15);
jurgan be yabu- “to perform one’s duty” (NL:166);
“For noble men (true gentlemen) to serve as high officials is to per
form (their) duties” (ORL:137);
amban serengge hese be ali-fi wen be high.official TOP edict ACC receive-CONV education ACC selgiye-re-ngge disseminate-PART-NR “High officials having received an imperial order, disseminate edu cation “ (PASH1:33). 8) In Manchu there are specific sentences, formally verbal, which are formed in accordance with the following syntactic pattern: S{Tv-ra-ngge} === P{Tv-mbi}. The new information (rheme) is rendered by the subject, not by the predicate, which is not obligatory from the semantic point of view. The subject is expressed by a nominalization. The nominalization rendering a subject and the verb rendering a predicate, are both formed from the same verbal stem: gabta-ra-ngge gabta-mbi geli shoot.an.arrow-PART-NR shoot-IMPF also gidala-ra-ngge gidala-mbi pierce.with.a.spear-PART-NR pierce-IMPF “Those who have to shoot, (they) shoot, and those who have to pierce, (they) pierce” (NSB:112). 9) Correlation can be realized by means of the syntactic pattern S {Tv-ra ba} === P {N}, where the subject is expressed by the form of the participle substantivized with the help of the substantivizer ba, and the predicate is rendered by a nominal word: manju bithe-de untuhun hergen baitala-ra ba umesi labdu manchu book-DAT empty letter use-PART SBSTR very many “In Manchu books auxiliary particles (the so-called empty words) are widely used” (PASH2: 51; JB). All syntactic patterns, listed above, manifest the syntactic relation ships they denote by means of the correlative forms of the subject and the predicate. However, many of them contain subjects ex pressed by predicative nominalizations which themselves semanti cally and structurally refer to a sentence. Strictly speaking, sentences
gorel-3.p65
435
4/17/02, 12:35 PM
436
part five
containing predicative nominalizations cannot be considered as simple. They function as predicative subjects in a sentence, syntac tically belonging to a higher level than that of a simple sentence and therefore should be discussed in the Syntax of complex sentences (polypredicative constructions) (see Section 5.4.3). 4.2. Grammatical Structure of a Verbal Sentence A verbal sentence renders a certain situation (event, process) of the outside world. Normally this situation includes some participants (persons, objects, etc.). Whether or not these participants are used in a sentence, the possibility of their presence is conditioned by the semantics of the verb. Each participant of a real situation is put into correlation with an actant. This term denotes the part of a sentence which designates a certain participant: an agent (a source of a ver bal action), or a patient (an object of a verbal action), or an experiencer (who perceives visual or auditory information), or a stimulus (a source of information for the experiencer), or an addressee (towards whom an action (or information) is referred), or a recipi ent (who becomes the possessor of the patient as a result of purposeful action of the agent), or a benefactive (who must benefit by the final result of an action), or an instrument (by which an action is per formed), or material (from which an object is made), or theme (what is spoken about). In some languages, certain participants, usually agents, may be rendered by a verbal form. In addition, a situation described by a sentence, possesses a number of characteristics which are not conditioned by the lexical meaning of the verb. Certain characteristics, very important for the meaning of an utterance, can be expressed by various adverbial words, which play circumstantial roles (< lat. circumstantia). Some characteristics are rendered by aspectual and temporal verbal forms. The ability of a verb to be combined with actants, i.e. the nomi nal parts of a sentence, and “to open places” for them, is called valency of a verb. Each verb possesses a certain number of valen cies which may determine the number of actants. Manchu verbs have not been thoroughly investigated from this point of view. Never theless, some old grammars contain information concerning the ability of verbs to combine with nouns which followed by certain case markers. Valency is considered obligatory when “a place” which a verb
gorel-3.p65
436
4/17/02, 12:35 PM
437
syntax
“opens” for a participant, is filled with a noun. Valency is consid ered optional when it may or may not be realized. Many Manchu verbs are oriented toward the subject (agent). This subject, however, owing to specific features of Manchu morphology, cannot be reflected in the verb form itself by means of predicative (or possessive) finite suffixes as is possible in many other Tungusic languages. This sub ject may either be expressed by an independent word form or it may not necessarily be expressed explicitly at all. In the latter case, the subject is reconstructed from the syntactic context. If a verb may have two or more actants one of them is hierarchi cally opposed to the other as a subject to objects. The subject is the principal actant, and grammatical superiority is determined by a verb. In Manchu, the subject of verbal sentences is expressed by the same forms as that of sentences with nominal predication. The subject can be rendered by nominalizations, for example: sin-i taci-ha-ngge labdu nonggibu-ha you(si/sin- SG)-GEN learn-PART-NR much advance-PART “You really have made very great progress” (MOLL:28) (literally: “Your studies have advanced very much”). However, the nominalizations are more typical of nominal sentences and more often take part in the formation of nominal predication. Besides, the predicative nominalizations, extended with subordinate words, in many cases do not meet the requirements of the simple sentence because the subject itself has a predicative structure. The structure of the predicative nominalizations more properly refers to that of a sentence. Complements are normally classified into direct and indirect. The parts of the sentence, which play circumstantial roles, are classified according to their meaning into adverbial modifiers of place, time, manner, cause, purpose, etc. It is yet to be resolved how to correctly classify an adverbial component that has, for example, the mean ing of place. In sentences such as “He lives in a village,” the com ponent “in a village” is normally considered to be an adverbial modifier of place. However, verbs are not used without such adver bial modifiers, which being obligatory components, may be consid ered as special actants. Traditionally, an attribute is also included in a sentence. How ever, an ordinary attribute is a member of a non-predicative word
gorel-3.p65
437
4/17/02, 12:35 PM
438
part five
combination, not a sentence. Such an attribute is not governed by a verb-predicate. By contrast, a predicative attribute is a component of a composite verbal-nominal predicate, and sentences containing such structural components, are a kind of contamination of verbal and nominal sentences: yasa tuwa-hai aldasi buce-mbi eye look-CONV short-lived die-IMPF
yasa tuwahai “in an instant,” “right before one’s eyes” (NL:313);
“In a split second (I) shall die young” (NSB:113).
4.2.1. Simple Verbal Predicate A simple verbal predicate is expressed by imperfect and perfect participles as well as finite verbal forms of the indicative and forms of oblique moods (imperative, optative, prohibitive). When serving as a simple verbal predicate, the imperfect parti ciple is occasionally used to denote the present, but more often it is used to denote the future: sain be tukiyece-rakå fas�s�a-ha be good ACC boast-PART(NEG) exert.effort-PART ACC bardanggila-rakå be buye-re brag-PART(NEG) ACC desire-PART
“(I) don’t want to boast of doing good or brag about my exertion”
(ORL:181);
bi sin-de emu sain arga tacibu-re
I you(si/sin- SG)-DAT one good method teach-PART
“I shall teach you a good method” (ORL:180).
Serving as a simple predicate, this form may be used in the mean
ing of the imperative or the optative when addressing people of equal
or lower social position (or status):
tuwa yata-ra fire strike.a.fire (with a flint) “Strike a fire;” min-i emu gusin be donji-re I(bi/min-)-GEN = my one word ACC listen-PART “Listen to one word of mine;”
gorel-3.p65
438
4/17/02, 12:35 PM
439
syntax bai-me gisu-re wish-CONV speak-PART “Tell me, please” (ORL:182).
The interrogative form of the imperfect participle in -rao/-reo may also serve as a simple predicate. Denoting the imperative, it is used to address older people or those of higher social position. Persons, addressed by the speaker, may be of equal age or social status, but in this case this form is used to express extreme politeness or hu mility. This meaning may be rendered by such expressions as “with your permission,” “by your leave,” “please, allow me to do smth.,” “I ask you humbly:” ala-ra-o tell-PART-INT “Please, tell me;” oncodo-ro-o forgive-PART-INT “Please, forgive me” (ZAKH:180); min-i jui i ajigen ergen be I(bi/min-)-GEN = my child GEN young life ACC aitubu-re-o resuscitate-PART-INT “Please, resuscitate the young life of my child” (NSB:121). When serving as a simple predicate the perfect participle usually denotes the past: deo bohori elan i ba-de te-he younger.brother PN place-name GEN place-DAT live-PART “(His) younger brother Bohori lived in Elan (place)” (GAB:91); julge-i ming gurun i forgon de emu ancient.times-GEN PN dynasty GEN time DAT one lolo se-re gasan bi-he place-name call-PART village be-PART “Many years ago, at the time of Ming dynasty there was a village called Lolo;” ambula urgunje-me gebu be uthai susai very.much rejoice-CONV name ACC thereupon fifty
gorel-3.p65
439
4/17/02, 12:35 PM
440
part five
se-de banji-ha year(of age)-DAT be.born-PART sergådai fiyanggå se-me gebule-fi PN PN(youngest) call-CONV name-CONV fiyanggå “youngest, smallest, least; the last-born child” (NL:89); “Having rejoiced greatly, (they) named him Sergådai Fiyanggå (be cause he) was born when they were both fifty” (NSB:111). The perfect participle may be followed by the sentence particle kai, which displaying affirmative modality, indicates the end of a sen tence: emgeri sin-i boo be taka-ha already you(si/sin- SG)-GEN = your house ACC recognize-PART kai MDL.PTL
“I have already found out where you live (your house)” (MOLL:38;
TM);
age yalu bi sin-de jaila-ha kai
sir ride(IMP) I you(si/sin- SG)-DAT avoide-PART COP
“Keep on your horse, sir, pray! I went out of your sight” (MOLL:47;
TM);
min-i haji sure jui emgeri buce-he
I(bi/min-)-GEN = my beloved wise child already die-PART
kai MDL.PTL “My beloved clever son has already died indeed” (NSB:115). The perfect participle may be followed by some other sentence particles, for example, the particle dere which denotes the modality of presumption: weihun gurun de uju tuci-ke amba gurun de alive country DAT first appear-PART big country DAT algin algi-ha nis�an saman fame be.famous-PART place-name shaman(ness) ji-fi gama-ha dere come-CONV take-PART MDL.PTL (probably)
“(I think) that Nishan shamaness who appeared in the counry of living
gorel-3.p65
440
4/17/02, 12:35 PM
441
syntax
beings and became famous in the great country, came and took (him)” (NSB:128). The forms of the indicative normally serve as simple verbal predi cates. The imperfect form in -mbi is mostly used to denote the present and the future tenses. It is occasionally used to denote the past tense, mostly when the form in -mbi is preceded by the perfect converb in -fi (for details see Part IV, Section 5.8.1). Here are some examples: gere-ndere de guwe-mbi become.bright-PART DAT sound-IMPF
“(Smth.) sounds at dawn” (literally: “When it becomes bright (smth.)
sounds”) (ORL:182);
ainu uttu gisure-mbi why.so/like.this talk-IMPF “What are you talking about!” (MOLL:19; TM); bithe hålana-me gene-mbi book go.to.read-CONV go-IMPF “(I) go to my studies” (MOLL:39; TM) (literally: “I go to read books”); holo saman holto-mbi lying shaman deceive-IMPF “A lying shaman deceives” (NSB:121); si ainu gele-ho golo-ho durun i ekàe-mbi you why fear-PART be.scared-PART shape GEN be.in.a.hurryIMPF “Why are you in a hurry seemingly so scared?” (NSB:118); cargi bakcin dalin de emu niyalma weihu be that.side the.opposite.side riverbank DAT one man boat ACC suru-me jabu-mbi punt.a.boat-CONV go-IMPF “From the opposite side of the riverbank a man is going punting a boat” (NSB:158); tookabi-ci suwem-be gemu tanta-mbi procrastinate-CONV you(suwe/suwen- PL)-ACC all beat-IMPF “If you procrastinate, (I) shall beat you all” (NSB:115); boo i gubci ere mejige be donji-fi gemu house GEN all this news ACC hear-CONV all songgoco-mbi cry-IMPF
gorel-3.p65
441
4/17/02, 12:35 PM
442
part five
“Having heard this news, all people of the household (gathered ) and cried” (NSB:114); ahalji bahalji geren ahasi sa kiyoo be àurde-me PN PN all servant PL litter ACC go.around-CONV uku-fi songgo-ro jilgan de alin crowd.around-CONV cry-PART voice DAT mountain holo gemu uradu-mbi valley all re-echo-IMPF “Ahalji, Bahalji, and all the servants crowded around the litter, and crying echoed though all the mountains and valleys (NSB:113). The form in -mbi may be followed by the sentence particle kai that serves as an additional sign denoting affirmative modality and indi cating the end of a sentence: jai uttu o-ho-de, bi usha-mbi-kai further like.this be-PART-DAT I be.angry-IMPF-MDL.PTL “If it goes on like this, I shall be angry” (ORL:193); min-i deo ere siden de teni I(bi/min-)-GEN) = my younger.brother this interval DAT highly hacihiya-me manju bithe håla-mbi-kai exert.oneself-CONV Manchu book read-IMPF-MDL.PTL “My younger brother is now working as hard as he can at Manchu” (MOLL:30; TM); solina-ci hono jide-rakå jalin invite-CONV yet come-PART(NEG) because.of joboào-mbi-kai be.greatly.distressed-IMPF-MDL.PTL
“I never asked you to come, because I feared you would refuse”
(MOLL:50; TM).
The form in -mbi may be followed by some other sentence particles,
for example, the particle dere which denotes the modality of
presumtion:
damu sin-i taci-rakå be
only you(si/sin- SG)-GEN = your learn-PART(NEG) ACC
hendu-mbi dere speak-IMPF MDL.PTL (probably)
“(My only fear would have been) that you were not anxious for
learning” (MOLL:19; TM);
gorel-3.p65
442
4/17/02, 12:35 PM
443
syntax
udu tere ten de isina-me mute-rakå although that hight DAT reach-CONV be.able-PART(NEG) bi-cibe inu urunakå be(AUX)-CONV certainly without.doubt hamina-mbi dere go.near.to-IMPF MDL.PTL (probably)
“Although we cannot reach the point he (has attained), we shall not
be very far behind him (I suspect)” (MOLL:25; TM).
Being a negative counterpart of the form in -mbi, the negative form
of the imperfect participle is normally used as the predicate of a
negative sentence:
doro erdemu gosin jurgan dorolon waka oci mute-rakå law wisdom mercy the.right ceremony NEG TOP be.achieved-
PART(NEG)
“Law, wisdom, mercy, right cannot be achieved without ceremony”
(ORL:193);
ama i gånin antaka be sa-rkå father GEN thought what.about ACC know-PART(NEG) “I would like to know what you, father, think about this?” (literally: “I don’t know my father’s thoughts (about this)” (NSB:111); ere yeken akå aha ainu ala-rakå this ignoble there.is.not servant why report-PART(NEG) damu songgo-mbi only cry-IMPF
yeken akå “ignoble;” “lowly,” “debased,” “not upright” (NL:314);
“Ignoble servant, why are you only crying and not reporting?”
(NSB:114);
tere sagda dosi-me jide-re de utala this old.man enter-CONV come-PART DAT so.much.as.this waliya-ra yali efen apki jergi be make.an.offering.at.a grave-PART meat bread wine et cetera ACC tuwa-rakå look-PART(NEG)
“When that old man came in, (he) didn’t (even) have a look at meat,
bread and wine” (NSB:117).
The indicative finite forms which denote the past, also function as
simple verbal predicates (for details see Part IV, Section 5.8.1):
gorel-3.p65
443
4/17/02, 12:35 PM
444
part five
1) the perfect form in -habi/-hebi/-hobi, -kabi/-kebi/-kobi, -ngkabi/
-ngkebi/-ngkobi:10
di se-me tukiye-hebi
emperor say-CONV call-PAST “(He) introduced (himself ) as the emperor;” abka-i fejergi be uherile-hebi heaven-GEN under ACC unite-PAST
abkai fejergi “all under heaven;”
“(He) united (all parts of) the empire together” (ORL:196);
jugån i andala nimeku baha-fi buce-habi road GEN halfway illness get-CONV die-PAST “Halfway he became ill and died” (NSB:111); tofohon se-de Heng Lang s�an alin fifteen year(of age)-DAT place-name mountain (chin.) mountain de abala-me gene-fi beye dube-hebi DAT hunt-CONV go-CONV himself die-PAST “When he was fifteen, he went to hunt on Henglang mountain and died” (NSB:111); emu amba omo de aisin menggun doohan ca-habi one big lake DAT gold silver bridge build(a bridge)-PAST “Across a big lake the bridge was built” (NSB:136); min-i beye nei taran tuci-kebi I(bi/min-)-GEN = my myself sweat (heavy)sweat appear-PAST nei taran “perspiration,” nei tuci- “to sweat” (NL:210); “I myself sweated heavily (being scared)” (NSB:130). The negative counterpart of this form, that is the form in -hakåbi/ -hekåbi, also serves as a simple verbal predicate. 2) the analytical form Tv-ha/-he/-ho bihe or Tv-ngka/-ngke/-ngko bihe: bi ergi emu niyamangga niyalma-i boo-de gene-he I this.side one related(by blood) man-GEN house-DAT go-PART bi-he be(AUX)-PART “I have been to visit (literally: to the house of) a relative of mine” (PASH2:172; TM). Mo� llendorff translated this as “I have been to visit a relation of mine who lives down yonder” (MOLL:35); 10 Permanent use of the copula bi after the perfect participle has led to the formation of the past form in -habi/-hebi/-hobi. The copula bi is often written separately from the participle bihe (bihe bi), but in other cases they are written as one word (bihebi). The unstable way of writing shows that the process of forming the past form had not been completed.
gorel-3.p65
444
4/17/02, 12:35 PM
445
syntax
sakda-si i tanggå se-de isina-ha erin-de old-PL GEN one.hundred year(of age) reach-PART time-DAT hiyoos�ula-me act.filially-CONV sinagala-me fude-ki seme majige gåni-ha bi-he mourn-CONV accompany-OPT that little think-PART be(AUX)PART “(I) thought to act filially when old (father and mother) reach their old age and (I) wanted mourning to accompany (them to their last journey)” (NSB:113). This form may be followed by the sentence particles bi or kai which are used to indicate the end of the speech. These particles may be written separately from or together with the component bihe (Zakharov, 1879:177). 3) the form in -mbihe: bi kemuni ere-be niyalma de ere-mbihe I constantly this-ACC people DAT hope-PAST “I constantly inspired people with this hope” (ORL:194). 4) the form in -mbihe may be followed by the sentence particle bi which is written either separately or as one word. This form is used to indicate usual and frequent actions which took place in the distant past: s�i irgebu-me mute-mbihebi verse compose-CONV be.able-PAST “(She) was able to compose verse (while still a child)” (ORL:194); waida-mbihe bi scoop.out-PAST COP “(He) used to scoop out (with a large spoon)” (ORL:78). 5) the analytical form Tv-me bihebi servs as a predicate. It is also used to indicate the distant past. Oriented towards the future, the optative form in -ki serves as a simple verbal predicate. Preserving its modal meaning of desire this form is normally used to denote the future tense of the indicative: muse uthai uttu o-ki we(INCL) then like.this act-OPT “We shall act like this” (ORL:173);
gorel-3.p65
445
4/17/02, 12:35 PM
446
part five
aina-ha seme baili be onggo-rakå urunakå do.what?-PART kindness ACC forget-PART(NEG) certainly ujele-me karula-ki value.highly-CONV repay-OPT ainaha seme “surely,” “without fail” (NL:9); “I shall never forget your kindness, and shall not fail to repay it handsomely” (MOLL:19; TM); encu inenggi jai cohome ji-fi gulhun emu inenggi other day again especially come-CONV entire one day gisure-me tece-ki speak-CONV sit.together-OPT
“I shall come another time and spend the day with you” (MOLL:39;
TM);
bi dosi-fi majige te-ki I enter-CONV a.little sit.down-OPT “I shall step in and sit down for a moment” (MOLL:48; TM). Finite forms of the oblique moods may also function as simple ver bal predicates (for grammatical data devoted to the oblique moods see Part IV, Sections 5.8.2, 5.8.2.1-3). The most widespread form for the imperative is that which coin cides with the verbal stem. Closely related to the honorific forms, this form denotes an address to the second person (singular and plural) occuping a lower or similar position in respect to the person who is speaking: cacari boo be sejen de tebu tent house ACC cart DAT put.in(IMP)
cacari boo “a square tent with cloth sides” (NL:40);
“Put a tent into a cart;”
niyalma morin enggemu jergi be teksile people horse saddle and.so.forth ACC prepare(IMP) “Prepare people, horses, and saddles;” cooha-i agåra beri niru jergi be belhe soldier-GEN weapon bow arrow and.so.forth ACC prepare(IMP)
coohai agåra “weapons” (NL:47);
“Prepare weapons, bows, and arrows” (NSB:112).
The meaning of order may be strengthened by using the personal
pronouns of the second person (singular and plural):
gorel-3.p65
446
4/17/02, 12:35 PM
447
syntax si majige ili you(SG) a.little stand(IMP)
“You stand still for a while” (ORL:166).
The imperative form may be followed by the syntactic words manggi
and nakå. However, predicates expressed by these combinations,
occur in dependent clauses with temporal meaning which can be
rendered by the conjunction ”after” (see Part IV, Section 8.3).
When the optative (form in -ki) refers to the second and third per sons (singular and plural) its meaning is very close to that of the imperative (see Part IV, Section 5.8.2.1): age wesi-fi te-ki elder.brother/sir ascend-CONV sit-OPT “Take the upper seat, please” (MOLL:36; TM); muse giyan be majige gisure-ki we(INCL) justice ACC a.little talk-OPT “Let us talk a little about fairness” (NSB:129). The imperative form in -kini, which is used to address a person who is lower in social position, may serve as a simple verbal predicate (see Part IV, Section 5.8.2.2): halhån o-ci majige tukiyecebu håwanggiyarakå hot be-CONV a.little take.away(IMP) it.makes.no.difference mukiyebu-kini cool.off-IMP “If it (tea) is too hot, let it be taken away (for a while), that it may get cooler” (MOLL:37; TM); ura s�urde-me uyun juse uji-kini buttocks spin-CONV nine children give.birth-IMP “Let (him) father nine children;” jalan as�s�a-me jakån jui sabu-kini generation move-CONV eight son see-IMP “Giving birth to generation let him see eight sons” (NSB:130). The imperative form in -cina/-kina may serve as a simple verbal predicate: majige je-fi gene-cina a.little eat-CONV go-IMP
gorel-3.p65
447
4/17/02, 12:35 PM
448
part five
“Do try and eat a little, then you may go” (MOLL:38; TM). The form of the prohibitive mood often functions as a simple ver bal predicate (for data concerning the prohibitive mood see Part IV, Section 5.8.2.3): nashån be ume ufara-bu-re opportunity ACC NEG miss-CAUS-PART “Do not throw away the opportunity” (MOLL:30; TM); jabs�a-bu-ci inu ume urgunje-re be.by.good.luck-PASS-CONV so NEG rejoice-PART ufara-bu-ci inu ume usaha-ra fail-PASS-CONV so NEG be.distressed-PART “If it succeeds by good luck, don’t rejoice, if it fails, don’t be dis tressed” (NSB:121). Interrogative verbal forms may also serve as a predicate in a simple sentence: age si manjura-me bahana-mbi-o elder.brother/sir speak.Manchu-CONV be.able-IMPF-INT “Sir, can you speak Manchu?” (PASH2: 216; DM); age sin-i boo-de manju sefu sir you(si/sin- SG)-GEN = your house-DAT manchu teacher soli-habi-o invite-PAST-INT “Sir, have you invited a teacher of the Manchu language (to give lessons on a regular basis) at your home?” (PASH2:223; DM); suwe hono in-i banin be you(PL) still/yet he(i/in-)-GEN = his nature ACC sa-rkå ni-o know-PART(NEG) INT-INT “Haven’t you known his character yet?” (PASH2:243; GH); age si ubade te-hebi-o sir you at.this.place/here live-PAST-INT “Sir, do you live here (at this place)?” (PASH2:172; TM). Mo�llendorff translated this as “Do you reside in this neighbourhood, sir?” (MOLL:36); sa-ha bi-ci aifini sim-be know-PART be(AUX)-CONV earlier you-ACC
gorel-3.p65
448
4/17/02, 12:35 PM
449
syntax
tuwanji-rakå bi-he-o come.to.see-PART(NEG) be(AUX)-PART-INT “If I only knew (that you lived here), I would have come to see you much earlier” (PAS